Release v0.1.16

.gitea/workflows/module-package-release.yml

@@ -0,0 +1,270 @@
+name: Module Package Release
+
+on:
+  push:
+    tags:
+      - "v*"
+  workflow_dispatch:
+    inputs:
+      release_tag:
+        description: Existing protected version tag to publish
+        required: true
+        type: string
+
+jobs:
+  publish-packages:
+    runs-on: ubuntu-latest
+    env:
+      GITEA_REPOSITORY: ${{ gitea.repository }}
+    steps:
+      - uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5
+        with:
+          fetch-depth: 0
+      - uses: actions/setup-python@a26af69be951a213d495a4c3e4e4022e16d87065
+        with:
+          python-version: "3.12"
+      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020
+        with:
+          node-version: "22"
+      - name: Select and validate protected release tag
+        shell: bash
+        env:
+          REQUESTED_TAG: ${{ inputs.release_tag }}
+          TRIGGER_TAG: ${{ gitea.ref_name }}
+        run: |
+          set -euo pipefail
+          tag="${REQUESTED_TAG:-$TRIGGER_TAG}"
+          case "$tag" in
+            v[0-9]*.[0-9]*.[0-9]*) ;;
+            *) echo "Release tag must start with a SemVer-shaped vX.Y.Z value" >&2; exit 1 ;;
+          esac
+          git fetch --force origin "refs/tags/$tag:refs/tags/$tag" refs/heads/main:refs/remotes/origin/main
+          tag_commit="$(git rev-list -n 1 "$tag")"
+          git merge-base --is-ancestor "$tag_commit" refs/remotes/origin/main || {
+            echo "Release tag is not contained in main" >&2
+            exit 1
+          }
+          git checkout --detach "$tag"
+          printf 'RELEASE_TAG=%s\n' "$tag" >> "$GITEA_ENV"
+          printf 'SOURCE_DATE_EPOCH=%s\n' "$(git show -s --format=%ct HEAD)" >> "$GITEA_ENV"
+      - name: Validate package versions
+        run: |
+          python - <<'PY'
+          import json
+          from pathlib import Path
+          import os
+          import re
+          import tomllib
+
+          tag = os.environ["RELEASE_TAG"]
+          expected = tag.removeprefix("v")
+          project = tomllib.loads(Path("pyproject.toml").read_text(encoding="utf-8"))["project"]
+          if project.get("version") != expected:
+              raise SystemExit(f"pyproject version {project.get('version')!r} does not match {tag}")
+          if re.fullmatch(r"govoplan-[a-z0-9-]+", str(project.get("name", ""))) is None:
+              raise SystemExit("Python distribution name must use the govoplan-* namespace")
+          webui = Path("webui/package.json")
+          if webui.is_file():
+              package = json.loads(webui.read_text(encoding="utf-8"))
+              if package.get("version") != expected:
+                  raise SystemExit(f"WebUI version {package.get('version')!r} does not match {tag}")
+              if re.fullmatch(r"@govoplan/[a-z0-9-]+-webui", str(package.get("name", ""))) is None:
+                  raise SystemExit("WebUI package name must use the @govoplan/*-webui namespace")
+              release = Path("webui/package.release.json")
+              if release.is_file():
+                  release_package = json.loads(release.read_text(encoding="utf-8"))
+                  if (
+                      release_package.get("name") != package.get("name")
+                      or release_package.get("version") != expected
+                  ):
+                      raise SystemExit("WebUI release package identity does not match package.json and the release tag")
+          PY
+      - name: Build immutable package artifacts
+        shell: bash
+        run: |
+          set -euo pipefail
+          python -m pip install --disable-pip-version-check build==1.5.0 twine==7.0.0
+          rm -rf dist .package-webui
+          python -m build --wheel --outdir dist
+          python -m twine check dist/*.whl
+          if [[ -f webui/package.json ]]; then
+            mkdir .package-webui
+            cp -a webui/. .package-webui/
+            rm -rf .package-webui/node_modules .package-webui/dist
+            if [[ -f .package-webui/package.release.json ]]; then
+              cp .package-webui/package.release.json .package-webui/package.json
+            fi
+          node <<'NODE'
+          const fs = require("node:fs");
+          const path = ".package-webui/package.json";
+          const packageJson = JSON.parse(fs.readFileSync(path, "utf8"));
+          const groups = ["dependencies", "optionalDependencies", "peerDependencies"];
+          for (const group of groups) {
+            for (const [name, specifier] of Object.entries(packageJson[group] || {})) {
+              if (!name.startsWith("@govoplan/")) continue;
+              if (typeof specifier !== "string") {
+                throw new Error(`${group}.${name} must use a string version`);
+              }
+              const packageSlug = name.slice("@govoplan/".length);
+              if (!packageSlug.endsWith("-webui")) {
+                throw new Error(`${group}.${name} is outside the WebUI package namespace`);
+              }
+              const repository = `govoplan-${packageSlug.slice(0, -"-webui".length)}`;
+              const escapedRepository = repository.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
+              const gitTag = specifier.match(
+                new RegExp(
+                  `^git\\+(?:ssh://git@|https://)git\\.add-ideas\\.de/(?:GovOPlaN|add-ideas)/${escapedRepository}\\.git#v([0-9]+\\.[0-9]+\\.[0-9]+)$`,
+                ),
+              );
+              if (gitTag) {
+                packageJson[group][name] = gitTag[1];
+                continue;
+              }
+              if (specifier.startsWith("file:") || specifier.startsWith("git+")) {
+                throw new Error(
+                  `${group}.${name} must resolve to an exact registry version for publication`,
+                );
+              }
+            }
+          }
+          delete packageJson.private;
+          fs.writeFileSync(path, `${JSON.stringify(packageJson, null, 2)}\n`);
+          NODE
+            npm pkg delete private --prefix .package-webui
+            (cd .package-webui && npm pack --ignore-scripts --pack-destination ../dist)
+          fi
+          python - <<'PY'
+          import hashlib
+          import json
+          from pathlib import Path
+          import os
+          import subprocess
+
+          artifacts = []
+          for path in sorted(Path("dist").iterdir()):
+              if path.suffix not in {".whl", ".tgz"}:
+                  continue
+              digest = hashlib.sha256(path.read_bytes()).hexdigest()
+              artifacts.append({"filename": path.name, "sha256": digest, "size": path.stat().st_size})
+          payload = {
+              "schema_version": "1",
+              "repository": os.environ["GITEA_REPOSITORY"],
+              "tag": os.environ["RELEASE_TAG"],
+              "commit": subprocess.check_output(["git", "rev-parse", "HEAD"], text=True).strip(),
+              "artifacts": artifacts,
+          }
+          Path("dist/package-artifacts.json").write_text(
+              json.dumps(payload, indent=2, sort_keys=True) + "\n",
+              encoding="utf-8",
+          )
+          PY
+      - name: Retain package hash evidence
+        uses: actions/upload-artifact@a8a3f3ad30e3422c9c7b888a15615d19a852ae32
+        with:
+          name: module-packages-${{ gitea.ref_name }}
+          path: dist/package-artifacts.json
+      - name: Check immutable registry state
+        shell: bash
+        env:
+          PACKAGE_TOKEN: ${{ secrets.GOVOPLAN_PACKAGE_TOKEN }}
+        run: |
+          set -euo pipefail
+          test -n "$PACKAGE_TOKEN"
+          python - <<'PY'
+          import hashlib
+          import json
+          import os
+          from pathlib import Path
+          import tomllib
+          from urllib.error import HTTPError
+          from urllib.parse import quote
+          from urllib.request import Request, urlopen
+
+          api_root = "https://git.add-ideas.de/api/v1/packages/GovOPlaN"
+          token = os.environ["PACKAGE_TOKEN"]
+
+          def should_publish(kind, name, version, path):
+              package_url = "/".join(
+                  (api_root, kind, quote(name, safe=""), quote(version, safe=""), "files")
+              )
+              request = Request(
+                  package_url,
+                  headers={"Accept": "application/json", "Authorization": f"token {token}"},
+              )
+              try:
+                  with urlopen(request, timeout=30) as response:
+                      files = json.load(response)
+              except HTTPError as exc:
+                  if exc.code == 404:
+                      print(f"{kind} package {name}=={version} is not published yet")
+                      return True
+                  raise
+              if not isinstance(files, list) or len(files) != 1:
+                  raise SystemExit(
+                      f"immutable {kind} package {name}=={version} has an unexpected file set"
+                  )
+              expected_sha256 = hashlib.sha256(path.read_bytes()).hexdigest()
+              if files[0].get("sha256") != expected_sha256:
+                  raise SystemExit(
+                      f"immutable {kind} package {name}=={version} already exists with a different SHA-256"
+                  )
+              print(f"verified existing {kind} package {name}=={version} ({expected_sha256})")
+              return False
+
+          project = tomllib.loads(Path("pyproject.toml").read_text(encoding="utf-8"))["project"]
+          wheels = tuple(Path("dist").glob("*.whl"))
+          if len(wheels) != 1:
+              raise SystemExit("release build must contain exactly one wheel")
+          publish_pypi = should_publish(
+              "pypi", str(project["name"]), str(project["version"]), wheels[0]
+          )
+
+          tarballs = tuple(Path("dist").glob("*.tgz"))
+          if len(tarballs) > 1:
+              raise SystemExit("release build must contain at most one npm package")
+          publish_npm = False
+          if tarballs:
+              webui = json.loads(
+                  Path(".package-webui/package.json").read_text(encoding="utf-8")
+              )
+              publish_npm = should_publish(
+                  "npm", str(webui["name"]), str(webui["version"]), tarballs[0]
+              )
+
+          with Path(os.environ["GITEA_ENV"]).open("a", encoding="utf-8") as env_file:
+              env_file.write(f"PUBLISH_PYPI={int(publish_pypi)}\n")
+              env_file.write(f"PUBLISH_NPM={int(publish_npm)}\n")
+          PY
+      - name: Publish wheel and WebUI package
+        shell: bash
+        env:
+          PACKAGE_USERNAME: ${{ secrets.GOVOPLAN_PACKAGE_USERNAME }}
+          PACKAGE_TOKEN: ${{ secrets.GOVOPLAN_PACKAGE_TOKEN }}
+        run: |
+          set -euo pipefail
+          test -n "$PACKAGE_USERNAME"
+          test -n "$PACKAGE_TOKEN"
+          if [[ "$PUBLISH_PYPI" == 1 ]]; then
+            TWINE_USERNAME="$PACKAGE_USERNAME" TWINE_PASSWORD="$PACKAGE_TOKEN" \
+              python -m twine upload --non-interactive \
+                --repository-url https://git.add-ideas.de/api/packages/GovOPlaN/pypi \
+                dist/*.whl
+          else
+            echo "Exact wheel is already present; skipping immutable retry."
+          fi
+          shopt -s nullglob
+          webui_packages=(dist/*.tgz)
+          if (( ${#webui_packages[@]} )) && [[ "$PUBLISH_NPM" == 1 ]]; then
+            npmrc="$(mktemp)"
+            trap 'rm -f "$npmrc"' EXIT
+            chmod 600 "$npmrc"
+            printf '%s\n' \
+              '@govoplan:registry=https://git.add-ideas.de/api/packages/GovOPlaN/npm/' \
+              "//git.add-ideas.de/api/packages/GovOPlaN/npm/:_authToken=$PACKAGE_TOKEN" \
+              > "$npmrc"
+            NPM_CONFIG_USERCONFIG="$npmrc" npm publish "./${webui_packages[0]}" \
+              --ignore-scripts --access public \
+              --registry https://git.add-ideas.de/api/packages/GovOPlaN/npm/
+          elif (( ${#webui_packages[@]} )); then
+            echo "Exact WebUI package is already present; skipping immutable retry."
+          fi

AGENTS.md

@@ -0,0 +1,16 @@
+# GovOPlaN Connectors Codex Guide
+
+## Scope
+
+This repository owns reusable external connection profiles, protocol adapters, governed snapshots, and connector capability contracts.
+
+## Documentation Contract
+
+- Treat documentation as part of every behavior change. Update this module's manifest-driven `DocumentationTopic` contributions for affected user and administrator behavior.
+- Keep feature content here; `govoplan-docs` projects it without importing Connectors internals.
+- Maintain a static user/admin baseline and run `/mnt/DATA/git/govoplan/tools/checks/check-manifest-shapes.py` after behavior or manifest changes.
+
+## Boundaries
+
+- Domain modules own business semantics; connectors own transport, credentials, synchronization, and diagnostics.
+- Keep optional adapters behind capabilities and enforce egress and peer-validation policy.

README.md

@@ -1,13 +1,58 @@
 # govoplan-connectors
 
-`govoplan-connectors` will own integration catalogues and generic external
+<!-- govoplan-repository-type:start -->
-system connection patterns for GovOPlaN.
+**Repository type:** connector (connector-hub).
+<!-- govoplan-repository-type:end -->
+
+`govoplan-connectors` owns integration catalogues and generic external system
+connection patterns for GovOPlaN.
 
 The module should make external systems discoverable, testable, and usable
 without taking ownership of their business semantics. Domain-specific modules
 remain responsible for case, file, workflow, payment, mail, identity, document,
 or reporting behavior.
 
+## Executable First Slice
+
+The first executable connector capability provides tenant-isolated tabular
+origins. Operators can import bounded JSON or CSV snapshots, inspect inferred
+schemas, and expose immutable source references and content fingerprints
+through `connectors.datasource_origins@0.1.0`. Preview reads enforce provider
+ceilings for rows, serialized bytes, and elapsed time and report the effective
+limits and any truncation as structured diagnostics.
+
+Connectors owns acquisition, connection profiles, credentials, discovery, and
+provider health. `govoplan-datasources` registers an origin as a governed live
+or cached datasource and owns staging, materializations, frozen states, and
+consumer access. Dataflow consumes that Datasources contract and never imports
+connector implementations or stores connector credentials.
+
+Each origin declares whether it is live, cached, file-backed, or static, its
+structured health state, and which projection, filter, aggregation, sorting,
+and pagination operations it can push down. The immutable snapshot provider
+currently supports projection and pagination only; consumers must keep other
+operations in Dataflow rather than assuming transport-side execution.
+
+Database, REST/HTTP, directory, managed-file, and warehouse providers can
+implement the same origin contract without changing Datasources or Dataflow.
+
+Governed sanctions and feed snapshot acquisitions use Core recovery operations.
+The source revision/cursor, redacted dry-run decision, canonical request digest,
+and distributed lease are durable before network I/O. Immutable snapshot rows
+and the terminal recovery checkpoint commit atomically, and an
+`Idempotency-Key` replays the committed result without contacting the provider.
+Current acquisition transports are read-only. The exported external-mutation
+recovery contract requires stable idempotency, provider verification, and
+operator reconciliation, but no connector currently claims a production write
+or delete path.
+
+Development:
+
+```bash
+/mnt/DATA/git/govoplan/.venv/bin/python -m pip install -e .
+/mnt/DATA/git/govoplan/.venv/bin/python -m unittest discover -s tests
+```
+
 See:
 
 - [Connector concept](docs/CONCEPT.md)

docs/CONCEPT.md

@@ -11,6 +11,13 @@ results, credential references, and generic integration events. Protocol-heavy
 or domain-heavy integrations may live in dedicated modules once their scope is
 clear.
 
+Connector capability is not source ownership. Each configured binding also
+declares whether GovOPlaN is authoritative, the external system is
+authoritative, GovOPlaN keeps a mirror, both sides use governed synchronization,
+GovOPlaN supplies only a governance overlay, or the object is link-only. The
+same connector may be configured differently by tenant, service, object type,
+or field group.
+
 Detailed follow-up documents:
 
 - [Public-sector integration catalogue](PUBLIC_SECTOR_INTEGRATION_CATALOGUE.md)
@@ -18,6 +25,20 @@ Detailed follow-up documents:
 - [OpenProject connector concept](OPENPROJECT_CONNECTOR.md)
 - [OpenDesk integration map](OPENDESK_INTEGRATION_MAP.md)
 
+## Shared runtime boundary
+
+Connector transports use the versioned Core runtime contract for bounded dry
+runs and diagnostics. Connectors owns endpoint discovery, authentication
+hand-off, protocol reads, retry/backoff, and source health. A consuming module
+owns domain mappings and mutations: Addresses, for example, owns contact and
+vCard semantics even when Connectors supplies reusable LDAP, CardDAV, Exchange,
+or Google transport patterns.
+
+Dry runs carry an immutable input hash, source revision and fingerprint,
+redacted effects, diagnostics, truncation state, and an apply token. Apply must
+reject a changed input or source revision. URLs and diagnostics never contain
+credential material; they retain only credential-envelope references.
+
 ## Ownership
 
 The module owns:
@@ -29,9 +50,13 @@ The module owns:
 - connector health status and last-test evidence
 - operator-visible integration inventory
 - cross-module discovery of available external capabilities
+- supported integration maturity, source-authority modes, operation limits,
+  and effect/reconciliation behavior for each connector type
 
 The module does not own:
 
+- governed datasource identity, staging, materializations, frozen states, or
+  consumer read semantics, owned by `govoplan-datasources`
 - file storage semantics, owned by files/DMS
 - identity provisioning semantics, owned by IDM/access
 - mail/calendar semantics, owned by mail/calendar
@@ -59,6 +84,9 @@ The module should integrate through:
 - module manifest metadata, route factories, permissions, and migrations
 - a connector catalogue API for listing available connector types
 - a connection profile API with secret references, not plaintext secrets
+- a provider declaration that composes authority mode, maturity, supported
+  operations, revisions/freshness, health, limits, idempotency, conflicts,
+  evidence, and reconciliation behavior
 - capability declarations such as `connectors.catalog`,
   `connectors.profileTester`, and `connectors.health`
 - events such as `connector.profile_created`, `connector.test_succeeded`,
@@ -69,6 +97,11 @@ Domain modules should ask whether a connector capability exists and request a
 profile/test result through core-mediated capabilities. They must not import
 connector implementation modules directly.
 
+Data-oriented consumers use a two-layer path: Connectors publishes a
+provider-specific datasource origin, then Datasources registers and governs it.
+Dataflow, Workflow, Reporting, and other consumers use Datasources rather than
+calling the connector origin directly.
+
 ## Reference Journeys
 
 ### OpenProject Connector First
@@ -106,6 +139,7 @@ The first implementation should provide:
 - WebUI catalogue and profile pages
 - configuration-package fragment support
 - generic external-reference DTOs
+- source-authority binding and provider-operation metadata
 - health summary provider
 
 ## Permissions

docs/CONNECTOR_SOURCE_LIFECYCLE.md

@@ -14,6 +14,22 @@ predictable and avoids hidden module imports.
 - `bidirectional`: GovOPlaN supports both directions with conflict detection and
   reconciliation rules.
 
+Direction describes transport. Every binding also needs a source-authority
+mode:
+
+- `native_authoritative`
+- `external_authoritative`
+- `external_mirror`
+- `governed_sync`
+- `governance_overlay`
+- `linked_reference`
+
+The authority mode and the connector's integration maturity are orthogonal. A
+bidirectional connector may be configured as an external mirror, and a native
+GovOPlaN object may publish to an external target without transferring
+authority. The effective binding must identify its scope and provenance rather
+than relying on a profile-wide `sync` boolean.
+
 ## Source Data Lifecycle
 
 Connector profiles have operational states, while individual external records
@@ -88,10 +104,12 @@ this lifecycle when a connector publishes status.
 2. Fetch only the minimal remote data required for the declared use case.
 3. Normalize into a connector-owned staging payload.
 4. Validate shape, required fields, and source trust level.
-5. Emit a core-mediated event such as `connector.record_discovered`.
+5. Publish data-shaped inputs as versioned datasource origins.
-6. Let domain modules claim or transform staged data through capabilities, not
+6. Let Datasources register live/cached origins or stage immutable snapshots.
-   imports.
+7. Let domain modules consume governed datasource references through
-7. Store external references with source system, object type, object id, version
+   capabilities, not imports.
+8. Emit a core-mediated event such as `connector.record_discovered`.
+9. Store external references with source system, object type, object id, version
    or ETag, and last-seen timestamp.
 
 ## Publish Flow
@@ -101,8 +119,11 @@ this lifecycle when a connector publishes status.
    shape.
 3. Connector sends the remote request.
 4. Connector stores the remote id, version/ETag, and response diagnostics.
-5. Connector emits `connector.record_published` or `connector.publish_failed`.
+5. A timeout or lost acknowledgement after dispatch becomes outcome-unknown,
-6. Domain module stores only the external-reference DTO and any domain result.
+   not an ordinary failure or permission to duplicate the command.
+6. Connector emits a confirmed, retryable, outcome-unknown, reconciled, or
+   corrected result event.
+7. Domain module stores only the external-reference DTO and any domain result.
 
 ## Reconciliation
 
@@ -114,6 +135,45 @@ Every connector that writes to an external system needs a reconciliation story:
 - retry policy for temporary failures
 - explicit operator action for destructive overwrite or deletion
 - audit trace from GovOPlaN record to external request and response summary
+- explicit requested, approved, dispatched, possibly-executed, confirmed, and
+  reconciled/corrected effect states
+
+## Durable recovery operations
+
+Connectors declares two recovery classes. A read-only acquisition into an
+immutable snapshot is `atomic`: the source revision or conditional cursor,
+redacted dry-run decision, canonical request digest, and distributed
+tenant/provider lease are durable before the fetch. The acquired domain rows
+and terminal Core recovery checkpoint commit in one PostgreSQL transaction. A
+caller-supplied `Idempotency-Key` replays that committed result without a second
+provider request. A failed or stale transaction has no remote mutation and may
+be repeated only as a new deliberate acquisition.
+
+An external create, update, publish, or delete is `forward_recovery`. It must
+start through the connector mutation recovery contract with a stable
+idempotency key, SHA-256 request digest, source revision/cursor, and dry-run
+evidence. Definitive rejection is terminal. A timeout or lost acknowledgement
+after dispatch is `outcome_unknown` and blocks replay until the owning connector
+verifies provider state. The contract and conformance tests exist; no current
+connector advertises a production external mutation, so write/delete adoption
+remains explicitly planned rather than implied.
+
+## Provider Declaration
+
+An executable connector type should publish machine-readable metadata for:
+
+- owned object and field groups, plus supported authority modes;
+- supported discovery, link, search, read, publish, synchronize, migrate, and
+  replacement maturity;
+- read/write/delete/preview/dry-run operations and bounded response limits;
+- revision/concurrency token, freshness, health, timeout, retry, and conflict
+  semantics;
+- idempotency and outcome-unknown handling;
+- evidence, rollback/compensation, correction, and reconciliation paths;
+- classification, purpose, retention, secret, degraded, and outage behavior.
+
+This declaration composes Core contracts. It does not move protocol behavior
+or domain semantics into Core or Connectors.
 
 ## Capability Boundary
 
@@ -124,6 +184,7 @@ should ask core for capabilities such as:
 - `connectors.profileTester`
 - `connectors.health`
 - `connectors.externalReferences`
+- `connectors.datasourceOrigins`
 - `connectors.sourceConsumer`
 - `connectors.sourcePublisher`
 
@@ -155,5 +216,6 @@ Before shipping an executable connector type:
 - Add unavailable-optional-module tests for every consuming domain module.
 - Add profile test and health status fixtures.
 - Add external-reference DTO tests.
+- Add source-authority and provider-declaration validation tests.
 - Add lifecycle transition tests for pause, retry, retirement, and uninstall
   guard behavior.

pyproject.toml

@@ -0,0 +1,25 @@
+[build-system]
+requires = ["setuptools>=69", "wheel"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "govoplan-connectors"
+version = "0.1.16"
+description = "Governed connector catalogue and tabular source capabilities for GovOPlaN."
+readme = "README.md"
+requires-python = ">=3.12"
+license = "AGPL-3.0-or-later"
+authors = [{ name = "GovOPlaN" }]
+dependencies = [
+  "defusedxml>=0.7,<1",
+  "govoplan-core>=0.1.16",
+]
+
+[tool.setuptools.packages.find]
+where = ["src"]
+
+[tool.setuptools.package-data]
+govoplan_connectors = ["py.typed"]
+
+[project.entry-points."govoplan.modules"]
+connectors = "govoplan_connectors.backend.manifest:get_manifest"

src/govoplan_connectors/__init__.py

@@ -0,0 +1,3 @@
+"""GovOPlaN Connectors module."""
+
+__all__: list[str] = []

src/govoplan_connectors/backend/__init__.py

@@ -0,0 +1 @@
+"""Connector backend package."""

src/govoplan_connectors/backend/datasource_origins.py

@@ -0,0 +1,147 @@
+from __future__ import annotations
+
+from govoplan_core.core.datasources import (
+    DatasourceAccessError,
+    DatasourceField,
+    DatasourceNotFoundError,
+    DatasourceOrigin,
+    DatasourceOriginReadRequest,
+    DatasourceOriginReadResult,
+    DatasourceUnavailableError,
+    DatasourceValidationError,
+)
+from govoplan_core.core.tabular_sources import (
+    TabularReadRequest,
+    TabularSource,
+    TabularSourceAccessError,
+    TabularSourceError,
+    TabularSourceNotFoundError,
+    TabularSourceUnavailableError,
+    TabularSourceValidationError,
+)
+from govoplan_connectors.backend.tabular_sources import SqlTabularSourceProvider
+
+
+class ConnectorDatasourceOriginProvider:
+    """Expose connector-owned sources through the Datasources origin contract."""
+
+    def __init__(self, provider: SqlTabularSourceProvider | None = None) -> None:
+        self._provider = provider or SqlTabularSourceProvider()
+
+    def list_origins(
+        self,
+        session: object,
+        principal: object,
+        *,
+        query: str = "",
+        limit: int = 100,
+    ):
+        try:
+            rows = self._provider.list_sources(
+                session,
+                principal,
+                query=query,
+                limit=limit,
+            )
+        except TabularSourceError as exc:
+            raise _datasource_error(exc) from exc
+        return tuple(_origin(source) for source in rows)
+
+    def get_origin(
+        self,
+        session: object,
+        principal: object,
+        *,
+        origin_ref: str,
+    ) -> DatasourceOrigin | None:
+        try:
+            source = self._provider.get_source(
+                session,
+                principal,
+                source_ref=origin_ref,
+            )
+        except TabularSourceError as exc:
+            raise _datasource_error(exc) from exc
+        return _origin(source) if source is not None else None
+
+    def read_origin(
+        self,
+        session: object,
+        principal: object,
+        *,
+        request: DatasourceOriginReadRequest,
+    ) -> DatasourceOriginReadResult:
+        try:
+            result = self._provider.read_source(
+                session,
+                principal,
+                request=TabularReadRequest(
+                    source_ref=request.origin_ref,
+                    limit=request.limit,
+                    offset=request.offset,
+                    columns=request.columns,
+                    expected_fingerprint=request.expected_fingerprint,
+                    max_bytes=request.max_bytes,
+                    timeout_ms=request.timeout_ms,
+                ),
+            )
+        except TabularSourceError as exc:
+            raise _datasource_error(exc) from exc
+        return DatasourceOriginReadResult(
+            origin=_origin(result.source),
+            rows=result.rows,
+            total_rows=result.total_rows,
+            truncated=result.truncated,
+            returned_bytes=result.returned_bytes,
+            elapsed_ms=result.elapsed_ms,
+            effective_row_limit=result.effective_row_limit,
+            effective_byte_limit=result.effective_byte_limit,
+            effective_timeout_ms=result.effective_timeout_ms,
+            diagnostics=result.diagnostics,
+        )
+
+
+def _origin(source: TabularSource) -> DatasourceOrigin:
+    return DatasourceOrigin(
+        ref=source.ref,
+        source_name=source.source_name,
+        name=source.name,
+        description=source.description,
+        kind="upload",
+        shape="tabular",
+        supported_modes=("live", "cached"),
+        provider=f"connectors.{source.provider}",
+        schema=tuple(
+            DatasourceField(
+                name=column.name,
+                data_type=column.data_type,
+                nullable=column.nullable,
+            )
+            for column in source.schema
+        ),
+        schema_version=source.schema_version,
+        fingerprint=source.fingerprint,
+        row_count=source.row_count,
+        byte_count=source.byte_count,
+        updated_at=source.updated_at,
+        capabilities=source.capabilities,
+        metadata=dict(source.metadata),
+        source_mode=source.source_mode,
+        pushdown=source.pushdown,
+        health=source.health,
+    )
+
+
+def _datasource_error(exc: TabularSourceError):
+    if isinstance(exc, TabularSourceAccessError):
+        return DatasourceAccessError(str(exc))
+    if isinstance(exc, TabularSourceNotFoundError):
+        return DatasourceNotFoundError(str(exc))
+    if isinstance(exc, TabularSourceUnavailableError):
+        return DatasourceUnavailableError(str(exc))
+    if isinstance(exc, TabularSourceValidationError):
+        return DatasourceValidationError(str(exc))
+    return DatasourceValidationError(str(exc))
+
+
+__all__ = ["ConnectorDatasourceOriginProvider"]

src/govoplan_connectors/backend/db/__init__.py

@@ -0,0 +1,3 @@
+from govoplan_connectors.backend.db.models import ConnectorTabularSource
+
+__all__ = ["ConnectorTabularSource"]

src/govoplan_connectors/backend/db/models.py

@@ -0,0 +1,263 @@
+from __future__ import annotations
+
+import uuid
+from datetime import datetime
+from typing import Any
+
+from sqlalchemy import (
+    DateTime,
+    ForeignKey,
+    Index,
+    Integer,
+    JSON,
+    LargeBinary,
+    String,
+    Text,
+    UniqueConstraint,
+)
+from sqlalchemy.orm import Mapped, mapped_column
+
+from govoplan_core.db.base import Base, TimestampMixin
+
+
+def new_uuid() -> str:
+    return str(uuid.uuid4())
+
+
+class ConnectorTabularSource(Base, TimestampMixin):
+    __tablename__ = "connector_tabular_sources"
+    __table_args__ = (
+        UniqueConstraint("tenant_id", "source_name", name="uq_connector_tabular_source_name"),
+        Index("ix_connector_tabular_sources_tenant_status", "tenant_id", "status"),
+        Index("ix_connector_tabular_sources_tenant_updated", "tenant_id", "updated_at"),
+    )
+
+    id: Mapped[str] = mapped_column(String(36), primary_key=True, default=new_uuid)
+    tenant_id: Mapped[str] = mapped_column(String(36), nullable=False, index=True)
+    provider: Mapped[str] = mapped_column(String(50), default="snapshot", nullable=False, index=True)
+    source_name: Mapped[str] = mapped_column(String(120), nullable=False)
+    name: Mapped[str] = mapped_column(String(300), nullable=False)
+    description: Mapped[str | None] = mapped_column(Text)
+    status: Mapped[str] = mapped_column(String(30), default="active", nullable=False, index=True)
+    schema_version: Mapped[int] = mapped_column(Integer, default=1, nullable=False)
+    schema_: Mapped[list[dict[str, Any]]] = mapped_column("schema", JSON, default=list, nullable=False)
+    rows: Mapped[list[dict[str, Any]]] = mapped_column(JSON, default=list, nullable=False)
+    fingerprint: Mapped[str] = mapped_column(String(64), nullable=False, index=True)
+    row_count: Mapped[int] = mapped_column(Integer, nullable=False)
+    byte_count: Mapped[int] = mapped_column(Integer, nullable=False)
+    metadata_: Mapped[dict[str, Any]] = mapped_column("metadata", JSON, default=dict, nullable=False)
+    created_by: Mapped[str | None] = mapped_column(String(255), nullable=True, index=True)
+    updated_by: Mapped[str | None] = mapped_column(String(255), nullable=True, index=True)
+    deleted_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True, index=True)
+
+
+class ConnectorSanctionsAcquisitionRun(Base, TimestampMixin):
+    __tablename__ = "connector_sanctions_acquisition_runs"
+    __table_args__ = (
+        Index(
+            "ix_connector_sanctions_run_health",
+            "tenant_id",
+            "provider_id",
+            "status",
+            "started_at",
+        ),
+    )
+
+    id: Mapped[str] = mapped_column(
+        String(36),
+        primary_key=True,
+        default=new_uuid,
+    )
+    tenant_id: Mapped[str] = mapped_column(
+        String(36),
+        nullable=False,
+        index=True,
+    )
+    provider_id: Mapped[str] = mapped_column(
+        String(100),
+        nullable=False,
+        index=True,
+    )
+    source_id: Mapped[str] = mapped_column(
+        String(200),
+        nullable=False,
+        index=True,
+    )
+    status: Mapped[str] = mapped_column(
+        String(40),
+        default="running",
+        nullable=False,
+        index=True,
+    )
+    attempt_count: Mapped[int] = mapped_column(
+        Integer,
+        default=0,
+        nullable=False,
+    )
+    request_evidence: Mapped[dict[str, Any]] = mapped_column(
+        JSON,
+        default=dict,
+        nullable=False,
+    )
+    response_evidence: Mapped[dict[str, Any]] = mapped_column(
+        JSON,
+        default=dict,
+        nullable=False,
+    )
+    started_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True),
+        nullable=False,
+        index=True,
+    )
+    finished_at: Mapped[datetime | None] = mapped_column(
+        DateTime(timezone=True),
+        nullable=True,
+    )
+    snapshot_id: Mapped[str | None] = mapped_column(
+        String(36),
+        nullable=True,
+        index=True,
+    )
+    error: Mapped[str | None] = mapped_column(
+        Text,
+        nullable=True,
+    )
+    created_by: Mapped[str | None] = mapped_column(
+        String(255),
+        nullable=True,
+        index=True,
+    )
+
+
+class ConnectorSanctionsSnapshot(Base, TimestampMixin):
+    __tablename__ = "connector_sanctions_snapshots"
+    __table_args__ = (
+        UniqueConstraint(
+            "connector_run_id",
+            name="uq_connector_sanctions_snapshot_run",
+        ),
+        Index(
+            "ix_connector_sanctions_snapshot_source",
+            "tenant_id",
+            "provider_id",
+            "acquired_at",
+        ),
+        Index(
+            "ix_connector_sanctions_snapshot_version",
+            "provider_id",
+            "source_id",
+            "source_version",
+        ),
+    )
+
+    id: Mapped[str] = mapped_column(
+        String(36),
+        primary_key=True,
+        default=new_uuid,
+    )
+    tenant_id: Mapped[str] = mapped_column(
+        String(36),
+        nullable=False,
+        index=True,
+    )
+    provider_id: Mapped[str] = mapped_column(
+        String(100),
+        nullable=False,
+        index=True,
+    )
+    publisher: Mapped[str] = mapped_column(
+        String(300),
+        nullable=False,
+    )
+    jurisdiction: Mapped[str] = mapped_column(
+        String(100),
+        nullable=False,
+        index=True,
+    )
+    list_type: Mapped[str] = mapped_column(
+        String(100),
+        nullable=False,
+        index=True,
+    )
+    source_id: Mapped[str] = mapped_column(
+        String(200),
+        nullable=False,
+        index=True,
+    )
+    source_version: Mapped[str] = mapped_column(
+        String(255),
+        nullable=False,
+        index=True,
+    )
+    publication_at: Mapped[datetime | None] = mapped_column(
+        DateTime(timezone=True),
+        nullable=True,
+    )
+    effective_at: Mapped[datetime | None] = mapped_column(
+        DateTime(timezone=True),
+        nullable=True,
+    )
+    acquired_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True),
+        nullable=False,
+        index=True,
+    )
+    source_url: Mapped[str | None] = mapped_column(
+        String(1500),
+        nullable=True,
+    )
+    content_type: Mapped[str] = mapped_column(
+        String(200),
+        nullable=False,
+    )
+    byte_count: Mapped[int] = mapped_column(
+        Integer,
+        nullable=False,
+    )
+    sha256: Mapped[str] = mapped_column(
+        String(64),
+        nullable=False,
+        index=True,
+    )
+    signature_evidence: Mapped[dict[str, Any]] = mapped_column(
+        JSON,
+        default=dict,
+        nullable=False,
+    )
+    parser_version: Mapped[str] = mapped_column(
+        String(100),
+        nullable=False,
+    )
+    licence_notes: Mapped[str | None] = mapped_column(
+        Text,
+        nullable=True,
+    )
+    trust_notes: Mapped[str | None] = mapped_column(
+        Text,
+        nullable=True,
+    )
+    connector_run_id: Mapped[str] = mapped_column(
+        ForeignKey(
+            "connector_sanctions_acquisition_runs.id",
+            ondelete="RESTRICT",
+        ),
+        nullable=False,
+        index=True,
+    )
+    transport_evidence: Mapped[dict[str, Any]] = mapped_column(
+        JSON,
+        default=dict,
+        nullable=False,
+    )
+    raw_content: Mapped[bytes] = mapped_column(
+        LargeBinary,
+        nullable=False,
+    )
+
+
+__all__ = [
+    "ConnectorSanctionsAcquisitionRun",
+    "ConnectorSanctionsSnapshot",
+    "ConnectorTabularSource",
+    "new_uuid",
+]

src/govoplan_connectors/backend/feeds.py

@@ -0,0 +1,413 @@
+from __future__ import annotations
+
+import hashlib
+import re
+import xml.etree.ElementTree as ET
+from collections.abc import Mapping
+from dataclasses import replace
+from datetime import datetime, timedelta, timezone
+from email.utils import format_datetime, parsedate_to_datetime
+
+from defusedxml import ElementTree as SafeET
+from defusedxml.common import DefusedXmlException
+
+from govoplan_core.core.feeds import (
+    FeedCapabilityError,
+    FeedDocument,
+    FeedEntry,
+    FeedProvider,
+    FeedRenderRequest,
+    FeedRenderResult,
+)
+from govoplan_core.security.http_fetch import fetch_http
+
+
+MAX_FEED_BYTES = 5_000_000
+ATOM_NS = "http://www.w3.org/2005/Atom"
+
+
+class ConnectorFeedProvider(FeedProvider):
+    def fetch(
+        self,
+        url: str,
+        *,
+        timeout: float = 15,
+        max_entries: int = 2_000,
+    ) -> FeedDocument:
+        try:
+            response = fetch_http(
+                url,
+                timeout=timeout,
+                label="RSS/Atom feed URL",
+                headers={
+                    "Accept": (
+                        "application/atom+xml, application/rss+xml, "
+                        "application/xml;q=0.9, text/xml;q=0.8"
+                    )
+                },
+                max_bytes=MAX_FEED_BYTES,
+            )
+        except Exception as exc:
+            raise FeedCapabilityError(f"Feed acquisition failed: {exc}") from exc
+        if response.status < 200 or response.status >= 300:
+            raise FeedCapabilityError(
+                f"Feed acquisition returned HTTP {response.status}."
+            )
+        content_type = _header(response.headers, "content-type")
+        document = self.parse(
+            response.body,
+            source_url=url,
+            content_type=content_type,
+            max_entries=max_entries,
+        )
+        acquired_at = datetime.now(timezone.utc)
+        return replace(
+            document,
+            acquired_at=acquired_at,
+            fresh_until=_fresh_until(response.headers, acquired_at),
+            etag=_header(response.headers, "etag"),
+            last_modified=_header(response.headers, "last-modified"),
+            metadata={
+                **dict(document.metadata),
+                "http_status": response.status,
+                "byte_count": len(response.body),
+            },
+        )
+
+    def parse(
+        self,
+        content: bytes,
+        *,
+        source_url: str,
+        content_type: str | None = None,
+        max_entries: int = 2_000,
+    ) -> FeedDocument:
+        if not content:
+            raise FeedCapabilityError("Feed content is empty.")
+        if len(content) > MAX_FEED_BYTES:
+            raise FeedCapabilityError(
+                f"Feeds are limited to {MAX_FEED_BYTES // 1_000_000} MB."
+            )
+        try:
+            root = SafeET.fromstring(content)
+        except (ET.ParseError, DefusedXmlException) as exc:
+            raise FeedCapabilityError(f"Feed XML is not safe or valid: {exc}") from exc
+        local_name = _local_name(root.tag)
+        if local_name == "rss":
+            document = _parse_rss(root, source_url=source_url, max_entries=max_entries)
+        elif local_name == "feed":
+            document = _parse_atom(root, source_url=source_url, max_entries=max_entries)
+        else:
+            raise FeedCapabilityError("The document is neither an RSS nor an Atom feed.")
+        return replace(
+            document,
+            content_type=content_type,
+            sha256=hashlib.sha256(content).hexdigest(),
+        )
+
+    def render(self, request: FeedRenderRequest) -> FeedRenderResult:
+        if not request.title.strip() or not request.feed_url.strip():
+            raise FeedCapabilityError("Feed title and feed URL are required.")
+        entries = tuple(
+            entry
+            for entry in request.entries
+            if entry.visibility in request.allowed_visibilities
+        )
+        root = (
+            _render_rss(request, entries)
+            if request.format == "rss"
+            else _render_atom(request, entries)
+        )
+        body = ET.tostring(root, encoding="utf-8", xml_declaration=True)
+        return FeedRenderResult(
+            format=request.format,
+            content_type=(
+                "application/rss+xml; charset=utf-8"
+                if request.format == "rss"
+                else "application/atom+xml; charset=utf-8"
+            ),
+            body=body,
+            included_entries=len(entries),
+            excluded_entries=len(request.entries) - len(entries),
+        )
+
+
+def feed_rows(document: FeedDocument) -> tuple[Mapping[str, object], ...]:
+    """Map feed entries to the connector tabular shape used by Datasources."""
+
+    return tuple(
+        {
+            "id": entry.id,
+            "title": entry.title,
+            "url": entry.url,
+            "summary": entry.summary,
+            "content": entry.content,
+            "author": entry.author,
+            "published_at": (
+                entry.published_at.isoformat() if entry.published_at else None
+            ),
+            "updated_at": entry.updated_at.isoformat() if entry.updated_at else None,
+            "categories": list(entry.categories),
+            "enclosures": [dict(item) for item in entry.enclosures],
+        }
+        for entry in document.entries
+    )
+
+
+def _parse_rss(root: ET.Element, *, source_url: str, max_entries: int) -> FeedDocument:
+    channel = _first_child(root, "channel")
+    if channel is None:
+        raise FeedCapabilityError("RSS feed is missing its channel element.")
+    entries: list[FeedEntry] = []
+    for item in _children(channel, "item"):
+        if len(entries) >= max_entries:
+            raise FeedCapabilityError(f"Feeds are limited to {max_entries:,} entries.")
+        url = _text(item, "link")
+        identifier = _text(item, "guid") or url or _entry_fallback_id(item)
+        entries.append(
+            FeedEntry(
+                id=identifier,
+                title=_text(item, "title") or "(Untitled)",
+                url=url,
+                summary=_text(item, "description"),
+                content=_text(item, "encoded"),
+                author=_text(item, "author") or _text(item, "creator"),
+                published_at=_parse_date(_text(item, "pubDate")),
+                categories=tuple(
+                    value for child in _children(item, "category")
+                    if (value := (child.text or "").strip())
+                ),
+                enclosures=tuple(
+                    {
+                        "url": child.attrib.get("url"),
+                        "media_type": child.attrib.get("type"),
+                        "size_bytes": _integer(child.attrib.get("length")),
+                    }
+                    for child in _children(item, "enclosure")
+                ),
+            )
+        )
+    return FeedDocument(
+        format="rss",
+        title=_text(channel, "title") or "Untitled feed",
+        source_url=source_url,
+        entries=tuple(entries),
+        description=_text(channel, "description"),
+        home_url=_text(channel, "link"),
+        language=_text(channel, "language"),
+        updated_at=_parse_date(
+            _text(channel, "lastBuildDate") or _text(channel, "pubDate")
+        ),
+    )
+
+
+def _parse_atom(root: ET.Element, *, source_url: str, max_entries: int) -> FeedDocument:
+    entries: list[FeedEntry] = []
+    for item in _children(root, "entry"):
+        if len(entries) >= max_entries:
+            raise FeedCapabilityError(f"Feeds are limited to {max_entries:,} entries.")
+        alternate = _atom_link(item, "alternate")
+        identifier = _text(item, "id") or alternate or _entry_fallback_id(item)
+        author = _first_child(item, "author")
+        entries.append(
+            FeedEntry(
+                id=identifier,
+                title=_text(item, "title") or "(Untitled)",
+                url=alternate,
+                summary=_text(item, "summary"),
+                content=_text(item, "content"),
+                author=_text(author, "name") if author is not None else None,
+                published_at=_parse_date(_text(item, "published")),
+                updated_at=_parse_date(_text(item, "updated")),
+                categories=tuple(
+                    value for child in _children(item, "category")
+                    if (value := (child.attrib.get("term") or "").strip())
+                ),
+                enclosures=tuple(
+                    {
+                        "url": child.attrib.get("href"),
+                        "media_type": child.attrib.get("type"),
+                        "size_bytes": _integer(child.attrib.get("length")),
+                    }
+                    for child in _children(item, "link")
+                    if child.attrib.get("rel") == "enclosure"
+                ),
+            )
+        )
+    return FeedDocument(
+        format="atom",
+        title=_text(root, "title") or "Untitled feed",
+        source_url=source_url,
+        entries=tuple(entries),
+        description=_text(root, "subtitle"),
+        home_url=_atom_link(root, "alternate"),
+        updated_at=_parse_date(_text(root, "updated")),
+    )
+
+
+def _render_rss(request: FeedRenderRequest, entries: tuple[FeedEntry, ...]) -> ET.Element:
+    ET.register_namespace("atom", ATOM_NS)
+    root = ET.Element("rss", {"version": "2.0"})
+    channel = ET.SubElement(root, "channel")
+    _element(channel, "title", request.title)
+    _element(channel, "link", request.home_url)
+    _element(channel, "description", request.description or request.title)
+    _element(channel, f"{{{ATOM_NS}}}link", None, {
+        "href": request.feed_url,
+        "rel": "self",
+        "type": "application/rss+xml",
+    })
+    if request.language:
+        _element(channel, "language", request.language)
+    for entry in entries:
+        item = ET.SubElement(channel, "item")
+        _element(item, "guid", entry.id, {"isPermaLink": "false"})
+        _element(item, "title", entry.title)
+        if entry.url:
+            _element(item, "link", entry.url)
+        if entry.summary or entry.content:
+            _element(item, "description", entry.summary or entry.content)
+        if entry.author:
+            _element(item, "author", entry.author)
+        date = entry.published_at or entry.updated_at
+        if date:
+            _element(item, "pubDate", format_datetime(_utc(date)))
+        for category in entry.categories:
+            _element(item, "category", category)
+        for enclosure in entry.enclosures:
+            attributes = {
+                "url": str(enclosure.get("url") or ""),
+                "type": str(enclosure.get("media_type") or "application/octet-stream"),
+                "length": str(enclosure.get("size_bytes") or 0),
+            }
+            if attributes["url"]:
+                _element(item, "enclosure", None, attributes)
+    return root
+
+
+def _render_atom(request: FeedRenderRequest, entries: tuple[FeedEntry, ...]) -> ET.Element:
+    ET.register_namespace("", ATOM_NS)
+    root = ET.Element(f"{{{ATOM_NS}}}feed")
+    _element(root, f"{{{ATOM_NS}}}id", request.feed_url)
+    _element(root, f"{{{ATOM_NS}}}title", request.title)
+    _element(root, f"{{{ATOM_NS}}}link", None, {"href": request.home_url})
+    _element(
+        root,
+        f"{{{ATOM_NS}}}link",
+        None,
+        {"href": request.feed_url, "rel": "self", "type": "application/atom+xml"},
+    )
+    latest = max(
+        (date for entry in entries for date in (entry.updated_at, entry.published_at) if date),
+        default=datetime.now(timezone.utc),
+    )
+    _element(root, f"{{{ATOM_NS}}}updated", _utc(latest).isoformat().replace("+00:00", "Z"))
+    if request.description:
+        _element(root, f"{{{ATOM_NS}}}subtitle", request.description)
+    for value in entries:
+        entry = ET.SubElement(root, f"{{{ATOM_NS}}}entry")
+        _element(entry, f"{{{ATOM_NS}}}id", value.id)
+        _element(entry, f"{{{ATOM_NS}}}title", value.title)
+        if value.url:
+            _element(entry, f"{{{ATOM_NS}}}link", None, {"href": value.url})
+        if value.summary:
+            _element(entry, f"{{{ATOM_NS}}}summary", value.summary)
+        if value.content:
+            _element(entry, f"{{{ATOM_NS}}}content", value.content, {"type": "html"})
+        updated = value.updated_at or value.published_at or latest
+        _element(entry, f"{{{ATOM_NS}}}updated", _utc(updated).isoformat().replace("+00:00", "Z"))
+        if value.published_at:
+            _element(entry, f"{{{ATOM_NS}}}published", _utc(value.published_at).isoformat().replace("+00:00", "Z"))
+        if value.author:
+            author = ET.SubElement(entry, f"{{{ATOM_NS}}}author")
+            _element(author, f"{{{ATOM_NS}}}name", value.author)
+        for category in value.categories:
+            _element(entry, f"{{{ATOM_NS}}}category", None, {"term": category})
+    return root
+
+
+def _children(element: ET.Element, name: str) -> tuple[ET.Element, ...]:
+    return tuple(child for child in element if _local_name(child.tag) == name)
+
+
+def _first_child(element: ET.Element, name: str) -> ET.Element | None:
+    return next((child for child in element if _local_name(child.tag) == name), None)
+
+
+def _text(element: ET.Element | None, name: str) -> str | None:
+    if element is None:
+        return None
+    child = _first_child(element, name)
+    if child is None:
+        return None
+    value = "".join(child.itertext()).strip()
+    return value or None
+
+
+def _local_name(tag: str) -> str:
+    return tag.rsplit("}", 1)[-1].split(":", 1)[-1]
+
+
+def _atom_link(element: ET.Element, relation: str) -> str | None:
+    for child in _children(element, "link"):
+        if (child.attrib.get("rel") or "alternate") == relation:
+            return child.attrib.get("href")
+    return None
+
+
+def _parse_date(value: str | None) -> datetime | None:
+    if not value:
+        return None
+    try:
+        parsed = parsedate_to_datetime(value)
+    except (TypeError, ValueError, OverflowError):
+        try:
+            parsed = datetime.fromisoformat(value.replace("Z", "+00:00"))
+        except ValueError:
+            return None
+    return _utc(parsed)
+
+
+def _utc(value: datetime) -> datetime:
+    if value.tzinfo is None:
+        return value.replace(tzinfo=timezone.utc)
+    return value.astimezone(timezone.utc)
+
+
+def _integer(value: str | None) -> int | None:
+    try:
+        return int(value) if value is not None else None
+    except ValueError:
+        return None
+
+
+def _entry_fallback_id(element: ET.Element) -> str:
+    body = ET.tostring(element, encoding="utf-8")
+    return f"urn:sha256:{hashlib.sha256(body).hexdigest()}"
+
+
+def _header(headers: Mapping[str, str], name: str) -> str | None:
+    lowered = name.casefold()
+    return next((value for key, value in headers.items() if key.casefold() == lowered), None)
+
+
+def _fresh_until(headers: Mapping[str, str], acquired_at: datetime) -> datetime | None:
+    cache_control = _header(headers, "cache-control") or ""
+    match = re.search(r"(?:^|,)\s*max-age\s*=\s*(\d+)", cache_control, re.IGNORECASE)
+    if match:
+        return acquired_at + timedelta(seconds=int(match.group(1)))
+    return _parse_date(_header(headers, "expires"))
+
+
+def _element(
+    parent: ET.Element,
+    tag: str,
+    text: str | None,
+    attributes: Mapping[str, str] | None = None,
+) -> ET.Element:
+    child = ET.SubElement(parent, tag, dict(attributes or {}))
+    child.text = text
+    return child
+
+
+__all__ = ["ConnectorFeedProvider", "MAX_FEED_BYTES", "feed_rows"]

src/govoplan_connectors/backend/manifest.py

@@ -0,0 +1,547 @@
+from __future__ import annotations
+
+from pathlib import Path
+
+from govoplan_core.core.access import (
+    CAPABILITY_AUTH_PERMISSION_EVALUATOR,
+    CAPABILITY_AUTH_PRINCIPAL_RESOLVER,
+)
+from govoplan_core.core.module_guards import (
+    drop_table_retirement_provider,
+    persistent_table_uninstall_guard,
+)
+from govoplan_core.core.datasources import CAPABILITY_DATASOURCE_ORIGINS
+from govoplan_core.core.feeds import CAPABILITY_CONNECTORS_FEEDS
+from govoplan_core.core.modules import (
+    DocumentationTopic,
+    MigrationSpec,
+    ModuleInterfaceProvider,
+    ModuleManifest,
+    PermissionDefinition,
+    RoleTemplate,
+)
+from govoplan_core.core.provider_governance import (
+    ExternalProviderDeclaration,
+    ExternalProviderStateProviderRegistration,
+    ModuleArchitectureDeclaration,
+    ModuleArchitectureDocumentation,
+    ModuleMaturityEvidence,
+    ProviderBehaviorDeclaration,
+    ProviderObjectDeclaration,
+)
+from govoplan_core.core.tabular_sources import (
+    CAPABILITY_CONNECTORS_TABULAR_SNAPSHOT_WRITER,
+    CAPABILITY_CONNECTORS_TABULAR_SOURCES,
+)
+from govoplan_core.core.sanctions import (
+    CAPABILITY_CONNECTORS_SANCTIONS_SNAPSHOTS,
+)
+from govoplan_core.db.base import Base
+from govoplan_connectors.backend.db.models import (
+    ConnectorSanctionsAcquisitionRun,
+    ConnectorSanctionsSnapshot,
+    ConnectorTabularSource,
+)
+from govoplan_connectors.backend.sanctions_sources import (
+    SANCTIONS_READ_SCOPE,
+    SANCTIONS_REFRESH_SCOPE,
+    SqlSanctionsSnapshotProvider,
+)
+from govoplan_connectors.backend.tabular_sources import (
+    ADMIN_SCOPE,
+    READ_SCOPE,
+    WRITE_SCOPE,
+    SqlTabularSourceProvider,
+)
+from govoplan_connectors.backend.datasource_origins import (
+    ConnectorDatasourceOriginProvider,
+)
+from govoplan_connectors.backend.feeds import ConnectorFeedProvider
+from govoplan_connectors.backend.provider_state import (
+    SANCTIONS_PROVIDER_ID,
+    TABULAR_PROVIDER_ID,
+    sanctions_provider_states,
+    tabular_provider_states,
+)
+
+
+MODULE_ID = "connectors"
+MODULE_VERSION = "0.1.16"
+TABULAR_SOURCE_INTERFACE_VERSION = "0.1.0"
+DATASOURCE_ORIGIN_INTERFACE_VERSION = "0.1.0"
+SANCTIONS_SNAPSHOT_INTERFACE_VERSION = "1.0.0"
+FEED_INTERFACE_VERSION = "0.1.0"
+CONNECTOR_RUNTIME_INTERFACE_VERSION = "1.0.0"
+
+ARCHITECTURE = ModuleArchitectureDeclaration(
+    layer="data_reporting_integration",
+    kind="integration",
+    maturity="vertical_slice",
+    evidence=(
+        ModuleMaturityEvidence(
+            kind="test",
+            reference="tests/test_tabular_sources.py",
+            summary="Exercises tenant-safe immutable tabular snapshots and bounded reads.",
+        ),
+        ModuleMaturityEvidence(
+            kind="test",
+            reference="tests/test_sanctions_sources.py",
+            summary="Exercises source acquisition health, checksums, retries, and immutable evidence.",
+        ),
+        ModuleMaturityEvidence(
+            kind="recovery",
+            reference="tests/test_recovery.py",
+            summary="Proves atomic snapshot commits, idempotent replay, distributed fences, tamper rejection, and unknown external-effect handling.",
+        ),
+        ModuleMaturityEvidence(
+            kind="documentation",
+            reference="docs/CONNECTOR_SOURCE_LIFECYCLE.md",
+            summary="Defines source lifecycle, authority, evidence, and outage boundaries.",
+        ),
+    ),
+    known_limits=(
+        "The executable generic datasource origin is an immutable tabular snapshot; database and arbitrary REST profiles remain future providers.",
+        "Feed publication renders a governed document but does not yet push it to an external publishing endpoint.",
+    ),
+    supported_authority_modes=(
+        "external_authoritative",
+        "external_mirror",
+        "linked_reference",
+    ),
+    owned_concepts=(
+        "external transport profiles",
+        "protocol interaction",
+        "immutable connector snapshots",
+        "connector acquisition health",
+    ),
+    non_owned_concepts=(
+        "datasource catalogue identity and lifecycle",
+        "domain records and business semantics",
+        "data transformations",
+        "screening dispositions",
+    ),
+    target_tested_providers=(
+        TABULAR_PROVIDER_ID,
+        SANCTIONS_PROVIDER_ID,
+    ),
+    documentation=ModuleArchitectureDocumentation(
+        migration=("src/govoplan_connectors/backend/migrations/versions",),
+        upgrade=("docs/CONNECTOR_SOURCE_LIFECYCLE.md",),
+        recovery=("docs/CONNECTOR_SOURCE_LIFECYCLE.md",),
+        security=("docs/CONNECTOR_SOURCE_LIFECYCLE.md",),
+        operations=("docs/CONNECTOR_SOURCE_LIFECYCLE.md",),
+    ),
+)
+
+EXTERNAL_PROVIDERS = (
+    ExternalProviderDeclaration(
+        id=TABULAR_PROVIDER_ID,
+        module_id=MODULE_ID,
+        label="Immutable tabular snapshot provider",
+        maturity="read",
+        operations=("discover", "search", "read", "preview", "dry_run"),
+        objects=(
+            ProviderObjectDeclaration(
+                object_type="tabular_source_snapshot",
+                field_groups=("identity", "schema", "rows", "source_provenance"),
+                authority_modes=("external_authoritative", "external_mirror"),
+                default_authority_mode="external_mirror",
+            ),
+        ),
+        behavior=ProviderBehaviorDeclaration(
+            revision_tokens="Source fingerprints and immutable snapshot ids are retained.",
+            concurrency="Reads may require the expected fingerprint; snapshots never mutate in place.",
+            freshness="Snapshot acquisition time and source timestamp are exposed.",
+            health="Import validation and source-read failures are explicit.",
+            max_read_items=1000,
+            idempotency="Feed imports accept a caller request key and replay the same committed immutable source without refetching.",
+            retry="Read-only acquisition may be retried only as a new deliberate request after a failed atomic operation.",
+            outcome_unknown="Provider reads do not mutate remote state; an uncertain database commit is resolved by the atomic recovery transaction.",
+            outcome_unknown_supported=False,
+            evidence="Rows, schema, fingerprint, source metadata, and acquisition provenance remain linked.",
+            correction="Import a replacement snapshot; retain the prior snapshot as evidence.",
+            rollback="Snapshot rows and the terminal recovery checkpoint commit or roll back together.",
+            reconciliation="Compare source and snapshot fingerprints before selecting a new current state.",
+            outage="Existing snapshots remain available and visibly stale; no live-source claim is made.",
+            classifications=("internal", "confidential", "restricted"),
+            purposes=("governed import", "dataflow input", "evidence reconstruction"),
+            retention="Datasources or the consuming domain supplies retention and hold policy.",
+            secret_handling="Generic snapshots contain no connector credential; transport credentials stay in credential envelopes.",
+        ),
+        capability_names=(
+            CAPABILITY_CONNECTORS_TABULAR_SOURCES,
+            CAPABILITY_DATASOURCE_ORIGINS,
+        ),
+        interface_names=(
+            "connectors.tabular_sources",
+            "connectors.datasource_origins",
+        ),
+        documentation_topic_ids=(
+            "connectors.authority-and-effects",
+            "connectors.tabular-sources",
+        ),
+    ),
+    ExternalProviderDeclaration(
+        id=SANCTIONS_PROVIDER_ID,
+        module_id=MODULE_ID,
+        label="Sanctions source snapshot provider",
+        maturity="read",
+        operations=("discover", "search", "read", "preview"),
+        objects=(
+            ProviderObjectDeclaration(
+                object_type="sanctions_source_snapshot",
+                field_groups=("source_identity", "raw_evidence", "entries", "acquisition_health"),
+                authority_modes=("external_authoritative", "external_mirror"),
+                default_authority_mode="external_mirror",
+            ),
+        ),
+        behavior=ProviderBehaviorDeclaration(
+            revision_tokens="Provider source version, ETag, Last-Modified, and SHA-256 digest are retained when available.",
+            concurrency="Refreshes use conditional source requests, a distributed per-tenant/provider fence, and immutable snapshots.",
+            freshness="Latest successful acquisition, source timestamp, and stale health are reported.",
+            health="Transport, parsing, source-change, and malformed-source states are explicit.",
+            max_read_items=5000,
+            idempotency="A caller request key identifies one acquisition run and replays its committed result without contacting the source again.",
+            retry="Bounded HTTP retries are safe because acquisition is read-only; failed runs require a new deliberate request key.",
+            timeout_seconds=30,
+            outcome_unknown="The external operation is read-only; snapshot rows and recovery evidence commit atomically.",
+            outcome_unknown_supported=False,
+            evidence="Raw source bytes, checksum, acquisition run, parser result, and normalized entry count are linked.",
+            correction="A corrected source creates a new immutable snapshot and acquisition run.",
+            rollback="A failed database transaction leaves no snapshot and the stale atomic fence resolves as failed.",
+            reconciliation="Compare source version and digest, then preserve both prior and corrected evidence.",
+            outage="The latest accepted snapshot stays usable with stale/unavailable source health.",
+            classifications=("public", "internal"),
+            purposes=("sanctions source acquisition", "compliance screening evidence"),
+            retention="Risk and Records policies determine accepted snapshot retention and legal holds.",
+            secret_handling="Public sources require no subject data or source credential; configured proxy secrets remain external to snapshots.",
+        ),
+        capability_names=(CAPABILITY_CONNECTORS_SANCTIONS_SNAPSHOTS,),
+        interface_names=("connectors.sanctions_snapshots",),
+        documentation_topic_ids=(
+            "connectors.authority-and-effects",
+            "connectors.sanctions-snapshots",
+        ),
+    ),
+)
+
+
+def _permission(scope: str, label: str, description: str) -> PermissionDefinition:
+    module_id, resource, action = scope.split(":", 2)
+    return PermissionDefinition(
+        scope=scope,
+        label=label,
+        description=description,
+        category="Connectors",
+        level="tenant",
+        module_id=module_id,
+        resource=resource,
+        action=action,
+    )
+
+
+PERMISSIONS = (
+    _permission(
+        READ_SCOPE,
+        "View tabular sources",
+        "Discover and preview policy-visible tabular connector sources.",
+    ),
+    _permission(
+        WRITE_SCOPE,
+        "Manage tabular sources",
+        "Import and retire bounded tabular snapshots.",
+    ),
+    _permission(
+        ADMIN_SCOPE,
+        "Administer connector sources",
+        "Manage every tenant connector source and future source policies.",
+    ),
+    _permission(
+        SANCTIONS_READ_SCOPE,
+        "View sanctions source evidence",
+        "Inspect immutable sanctions snapshots and acquisition health.",
+    ),
+    _permission(
+        SANCTIONS_REFRESH_SCOPE,
+        "Refresh sanctions sources",
+        "Acquire a new immutable sanctions source snapshot.",
+    ),
+)
+
+ROLE_TEMPLATES = (
+    RoleTemplate(
+        slug="connector_source_manager",
+        name="Connector source manager",
+        description="Discover, import, preview, and retire tabular sources.",
+        permissions=(
+            READ_SCOPE,
+            WRITE_SCOPE,
+            SANCTIONS_READ_SCOPE,
+            SANCTIONS_REFRESH_SCOPE,
+        ),
+    ),
+    RoleTemplate(
+        slug="connector_source_reader",
+        name="Connector source reader",
+        description="Discover and preview tabular connector sources.",
+        permissions=(READ_SCOPE, SANCTIONS_READ_SCOPE),
+    ),
+)
+
+
+def _router(_context):
+    from govoplan_connectors.backend.router import router
+
+    return router
+
+
+def _provider(_context) -> SqlTabularSourceProvider:
+    return SqlTabularSourceProvider()
+
+
+def _datasource_origin_provider(_context) -> ConnectorDatasourceOriginProvider:
+    return ConnectorDatasourceOriginProvider()
+
+
+def _sanctions_snapshot_provider(
+    _context,
+) -> SqlSanctionsSnapshotProvider:
+    return SqlSanctionsSnapshotProvider()
+
+
+def _feed_provider(_context) -> ConnectorFeedProvider:
+    return ConnectorFeedProvider()
+
+
+def _tenant_summary(session, tenant_id: str) -> dict[str, int]:
+    return {
+        "connector_tabular_sources": (
+            session.query(ConnectorTabularSource)
+            .filter(
+                ConnectorTabularSource.tenant_id == tenant_id,
+                ConnectorTabularSource.deleted_at.is_(None),
+            )
+            .count()
+        ),
+        "connector_sanctions_snapshots": (
+            session.query(ConnectorSanctionsSnapshot)
+            .filter(ConnectorSanctionsSnapshot.tenant_id == tenant_id)
+            .count()
+        ),
+        "connector_sanctions_runs": (
+            session.query(ConnectorSanctionsAcquisitionRun)
+            .filter(ConnectorSanctionsAcquisitionRun.tenant_id == tenant_id)
+            .count()
+        ),
+    }
+
+
+manifest = ModuleManifest(
+    id=MODULE_ID,
+    name="Connectors",
+    version=MODULE_VERSION,
+    optional_dependencies=(
+        "access",
+        "audit",
+        "files",
+        "policy",
+        "datasources",
+        "portal",
+        "reporting",
+        "risk_compliance",
+    ),
+    required_capabilities=(
+        CAPABILITY_AUTH_PRINCIPAL_RESOLVER,
+        CAPABILITY_AUTH_PERMISSION_EVALUATOR,
+    ),
+    provides_interfaces=(
+        ModuleInterfaceProvider(
+            name="connectors.tabular_sources",
+            version=TABULAR_SOURCE_INTERFACE_VERSION,
+        ),
+        ModuleInterfaceProvider(
+            name="connectors.tabular_snapshot_writer",
+            version=TABULAR_SOURCE_INTERFACE_VERSION,
+        ),
+        ModuleInterfaceProvider(
+            name="connectors.datasource_origins",
+            version=DATASOURCE_ORIGIN_INTERFACE_VERSION,
+        ),
+        ModuleInterfaceProvider(
+            name="connectors.sanctions_snapshots",
+            version=SANCTIONS_SNAPSHOT_INTERFACE_VERSION,
+        ),
+        ModuleInterfaceProvider(
+            name="connectors.feeds",
+            version=FEED_INTERFACE_VERSION,
+        ),
+        ModuleInterfaceProvider(
+            name="connectors.runtime_contract",
+            version=CONNECTOR_RUNTIME_INTERFACE_VERSION,
+        ),
+    ),
+    permissions=PERMISSIONS,
+    role_templates=ROLE_TEMPLATES,
+    route_factory=_router,
+    capability_factories={
+        CAPABILITY_CONNECTORS_TABULAR_SOURCES: _provider,
+        CAPABILITY_CONNECTORS_TABULAR_SNAPSHOT_WRITER: _provider,
+        CAPABILITY_DATASOURCE_ORIGINS: _datasource_origin_provider,
+        CAPABILITY_CONNECTORS_SANCTIONS_SNAPSHOTS: (_sanctions_snapshot_provider),
+        CAPABILITY_CONNECTORS_FEEDS: _feed_provider,
+    },
+    tenant_summary_providers=(_tenant_summary,),
+    architecture=ARCHITECTURE,
+    external_providers=EXTERNAL_PROVIDERS,
+    external_provider_state_providers=(
+        ExternalProviderStateProviderRegistration(
+            module_id=MODULE_ID,
+            provider_id=TABULAR_PROVIDER_ID,
+            provider=tabular_provider_states,
+        ),
+        ExternalProviderStateProviderRegistration(
+            module_id=MODULE_ID,
+            provider_id=SANCTIONS_PROVIDER_ID,
+            provider=sanctions_provider_states,
+        ),
+    ),
+    migration_spec=MigrationSpec(
+        module_id=MODULE_ID,
+        metadata=Base.metadata,
+        script_location=str(Path(__file__).with_name("migrations") / "versions"),
+        retirement_supported=True,
+        retirement_provider=drop_table_retirement_provider(
+            ConnectorSanctionsSnapshot,
+            ConnectorSanctionsAcquisitionRun,
+            ConnectorTabularSource,
+            label="Connectors",
+        ),
+        retirement_notes=(
+            "Destructive retirement drops connector-owned source snapshots after "
+            "the installer captures a database snapshot."
+        ),
+    ),
+    uninstall_guard_providers=(
+        persistent_table_uninstall_guard(
+            ConnectorSanctionsSnapshot,
+            ConnectorSanctionsAcquisitionRun,
+            ConnectorTabularSource,
+            label="Connectors",
+        ),
+    ),
+    documentation=(
+        DocumentationTopic(
+            id="connectors.authority-and-effects",
+            title="Connector authority and effect behavior",
+            summary="Connector direction, technical maturity, and configured source authority are separate and must remain visible.",
+            body=(
+                "A connector can consume, publish, or work bidirectionally and can mature from discovery through replacement. "
+                "Each binding separately states whether GovOPlaN is authoritative, follows an external authority, keeps a mirror, synchronizes under conflict rules, adds a governance overlay, or retains only a link. "
+                "Writable providers must explain revisions, limits, idempotency, outcome-unknown handling, evidence, reconciliation, correction, outage behavior, and secret requirements."
+            ),
+            layer="available",
+            documentation_types=("admin", "user"),
+            audience=("operator", "module_admin", "power_user", "product_owner"),
+            related_modules=("datasources", "dataflow", "ops", "policy", "audit"),
+            order=39,
+        ),
+        DocumentationTopic(
+            id="connectors.runtime-preview-contract",
+            title="Connector previews and diagnostics",
+            summary="Use one bounded, redacted dry-run shape across external transports.",
+            body=(
+                "Connectors owns endpoint discovery, authentication hand-off, transport limits, retries, and protocol health. "
+                "Domain modules own field mapping, validation, reconciliation, and record mutation. The shared Core runtime "
+                "contract reports redacted effects and diagnostics with source revisions, fingerprints, and immutable input hashes. "
+                "Tabular previews enforce effective row, serialized-byte, and elapsed-time ceilings and report limit truncation "
+                "as structured diagnostics. A commit must reject stale, truncated, conflicting, or error-bearing previews, and "
+                "credentials never appear in URLs or samples."
+            ),
+            layer="available",
+            documentation_types=("admin", "user"),
+            audience=("operator", "module_admin", "power_user"),
+            related_modules=("addresses", "datasources", "dataflow", "policy", "audit"),
+            order=40,
+        ),
+        DocumentationTopic(
+            id="connectors.tabular-sources",
+            title="Governed tabular sources",
+            summary="Provider-neutral source discovery and bounded reads for Dataflow.",
+            body=(
+                "Connectors owns source configuration, access checks, schema discovery, "
+                "fingerprints, and bounded reads. Dataflow stores only opaque source "
+                "references and expected fingerprints. Each source declares its live, "
+                "cached, file-backed, or static mode, structured health, and supported "
+                "projection, filter, aggregation, sorting, and pagination pushdown. The "
+                "first executable provider imports immutable JSON or CSV snapshots, "
+                "supports projection and pagination, and exposes them as Datasource "
+                "origins. Database and API providers can implement the same origin "
+                "contract without changing Datasources or Dataflow."
+            ),
+            layer="available",
+            documentation_types=("admin", "user"),
+            audience=("operator", "module_admin", "power_user"),
+            related_modules=("dataflow", "files", "reporting", "risk_compliance"),
+            order=40,
+        ),
+        DocumentationTopic(
+            id="connectors.rss-atom",
+            title="RSS and Atom feeds",
+            summary="Import governed feed snapshots and emit visibility-filtered feeds.",
+            body=(
+                "Connectors owns bounded, SSRF-protected RSS/Atom transport and XML "
+                "parsing. Imported entries become immutable tabular snapshots exposed "
+                "through Datasources, including acquisition, freshness, ETag, content "
+                "digest, and source provenance. Portal or Reporting owns publication "
+                "routes and must pass the allowed visibility set when rendering output. "
+                "A separate RSS module is only warranted if GovOPlaN later needs a "
+                "dedicated feed-reader product surface."
+            ),
+            layer="available",
+            documentation_types=("admin", "user"),
+            audience=("operator", "module_admin", "power_user"),
+            related_modules=("datasources", "dataflow", "portal", "reporting"),
+            order=42,
+        ),
+        DocumentationTopic(
+            id="connectors.sanctions-snapshots",
+            title="Sanctions source snapshots",
+            summary=(
+                "Acquire immutable, checksum-verifiable sanctions list "
+                "evidence without transmitting screening subjects."
+            ),
+            body=(
+                "Connectors provides a deterministic synthetic fixture and "
+                "the official United Nations Security Council consolidated "
+                "XML source. Each fetch records conditional transport "
+                "evidence, bounded retries, health state, source metadata, "
+                "raw evidence, and a SHA-256 checksum. Refreshes acquire a "
+                "distributed recovery fence before provider I/O; the immutable "
+                "snapshot and terminal recovery checkpoint then commit in one "
+                "transaction. A repeated request key returns the same result. "
+                "Risk Compliance owns "
+                "normalization, matching, legal review, and dispositions."
+            ),
+            layer="available",
+            documentation_types=("admin", "user"),
+            audience=("operator", "module_admin", "compliance_reviewer"),
+            related_modules=("risk_compliance", "dataflow"),
+            order=41,
+        ),
+    ),
+)
+
+
+def get_manifest() -> ModuleManifest:
+    return manifest
+
+
+__all__ = [
+    "MODULE_ID",
+    "MODULE_VERSION",
+    "DATASOURCE_ORIGIN_INTERFACE_VERSION",
+    "SANCTIONS_SNAPSHOT_INTERFACE_VERSION",
+    "TABULAR_SOURCE_INTERFACE_VERSION",
+    "get_manifest",
+    "manifest",
+]

src/govoplan_connectors/backend/migrations/__init__.py

@@ -0,0 +1 @@
+"""Connectors migrations."""

src/govoplan_connectors/backend/migrations/versions/__init__.py

@@ -0,0 +1 @@
+"""Connectors migration revisions."""

src/govoplan_connectors/backend/migrations/versions/e6b7c8d9f0a1_v0114_connectors_baseline.py

@@ -0,0 +1,141 @@
+"""v0.1.14 Connectors baseline
+
+Revision ID: e6b7c8d9f0a1
+Revises: None
+Create Date: 2026-07-28 00:00:00.000000
+"""
+from __future__ import annotations
+
+from alembic import op
+import sqlalchemy as sa
+
+
+revision = "e6b7c8d9f0a1"
+down_revision = None
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.create_table(
+        "connector_tabular_sources",
+        sa.Column("id", sa.String(length=36), nullable=False),
+        sa.Column("tenant_id", sa.String(length=36), nullable=False),
+        sa.Column("provider", sa.String(length=50), nullable=False),
+        sa.Column("source_name", sa.String(length=120), nullable=False),
+        sa.Column("name", sa.String(length=300), nullable=False),
+        sa.Column("description", sa.Text(), nullable=True),
+        sa.Column("status", sa.String(length=30), nullable=False),
+        sa.Column("schema_version", sa.Integer(), nullable=False),
+        sa.Column("schema", sa.JSON(), nullable=False),
+        sa.Column("rows", sa.JSON(), nullable=False),
+        sa.Column("fingerprint", sa.String(length=64), nullable=False),
+        sa.Column("row_count", sa.Integer(), nullable=False),
+        sa.Column("byte_count", sa.Integer(), nullable=False),
+        sa.Column("metadata", sa.JSON(), nullable=False),
+        sa.Column("created_by", sa.String(length=255), nullable=True),
+        sa.Column("updated_by", sa.String(length=255), nullable=True),
+        sa.Column("deleted_at", sa.DateTime(timezone=True), nullable=True),
+        sa.Column("created_at", sa.DateTime(timezone=True), nullable=False),
+        sa.Column("updated_at", sa.DateTime(timezone=True), nullable=False),
+        sa.PrimaryKeyConstraint("id", name=op.f("pk_connector_tabular_sources")),
+        sa.UniqueConstraint(
+            "tenant_id",
+            "source_name",
+            name="uq_connector_tabular_source_name",
+        ),
+    )
+    op.create_index(
+        op.f("ix_connector_tabular_sources_created_by"),
+        "connector_tabular_sources",
+        ["created_by"],
+        unique=False,
+    )
+    op.create_index(
+        op.f("ix_connector_tabular_sources_deleted_at"),
+        "connector_tabular_sources",
+        ["deleted_at"],
+        unique=False,
+    )
+    op.create_index(
+        op.f("ix_connector_tabular_sources_fingerprint"),
+        "connector_tabular_sources",
+        ["fingerprint"],
+        unique=False,
+    )
+    op.create_index(
+        op.f("ix_connector_tabular_sources_provider"),
+        "connector_tabular_sources",
+        ["provider"],
+        unique=False,
+    )
+    op.create_index(
+        op.f("ix_connector_tabular_sources_status"),
+        "connector_tabular_sources",
+        ["status"],
+        unique=False,
+    )
+    op.create_index(
+        op.f("ix_connector_tabular_sources_tenant_id"),
+        "connector_tabular_sources",
+        ["tenant_id"],
+        unique=False,
+    )
+    op.create_index(
+        op.f("ix_connector_tabular_sources_updated_by"),
+        "connector_tabular_sources",
+        ["updated_by"],
+        unique=False,
+    )
+    op.create_index(
+        "ix_connector_tabular_sources_tenant_status",
+        "connector_tabular_sources",
+        ["tenant_id", "status"],
+        unique=False,
+    )
+    op.create_index(
+        "ix_connector_tabular_sources_tenant_updated",
+        "connector_tabular_sources",
+        ["tenant_id", "updated_at"],
+        unique=False,
+    )
+
+
+def downgrade() -> None:
+    op.drop_index(
+        "ix_connector_tabular_sources_tenant_updated",
+        table_name="connector_tabular_sources",
+    )
+    op.drop_index(
+        "ix_connector_tabular_sources_tenant_status",
+        table_name="connector_tabular_sources",
+    )
+    op.drop_index(
+        op.f("ix_connector_tabular_sources_updated_by"),
+        table_name="connector_tabular_sources",
+    )
+    op.drop_index(
+        op.f("ix_connector_tabular_sources_tenant_id"),
+        table_name="connector_tabular_sources",
+    )
+    op.drop_index(
+        op.f("ix_connector_tabular_sources_status"),
+        table_name="connector_tabular_sources",
+    )
+    op.drop_index(
+        op.f("ix_connector_tabular_sources_provider"),
+        table_name="connector_tabular_sources",
+    )
+    op.drop_index(
+        op.f("ix_connector_tabular_sources_fingerprint"),
+        table_name="connector_tabular_sources",
+    )
+    op.drop_index(
+        op.f("ix_connector_tabular_sources_deleted_at"),
+        table_name="connector_tabular_sources",
+    )
+    op.drop_index(
+        op.f("ix_connector_tabular_sources_created_by"),
+        table_name="connector_tabular_sources",
+    )
+    op.drop_table("connector_tabular_sources")

src/govoplan_connectors/backend/migrations/versions/f7c8d9e0a1b2_v0114_sanctions_snapshots.py

@@ -0,0 +1,192 @@
+"""Add immutable sanctions source snapshots.
+
+Revision ID: f7c8d9e0a1b2
+Revises: e6b7c8d9f0a1
+Create Date: 2026-07-29
+"""
+
+from __future__ import annotations
+
+from alembic import op
+import sqlalchemy as sa
+
+
+revision = "f7c8d9e0a1b2"
+down_revision = "e6b7c8d9f0a1"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.create_table(
+        "connector_sanctions_acquisition_runs",
+        sa.Column("id", sa.String(length=36), nullable=False),
+        sa.Column("tenant_id", sa.String(length=36), nullable=False),
+        sa.Column("provider_id", sa.String(length=100), nullable=False),
+        sa.Column("source_id", sa.String(length=200), nullable=False),
+        sa.Column("status", sa.String(length=40), nullable=False),
+        sa.Column("attempt_count", sa.Integer(), nullable=False),
+        sa.Column("request_evidence", sa.JSON(), nullable=False),
+        sa.Column("response_evidence", sa.JSON(), nullable=False),
+        sa.Column(
+            "started_at",
+            sa.DateTime(timezone=True),
+            nullable=False,
+        ),
+        sa.Column(
+            "finished_at",
+            sa.DateTime(timezone=True),
+            nullable=True,
+        ),
+        sa.Column("snapshot_id", sa.String(length=36), nullable=True),
+        sa.Column("error", sa.Text(), nullable=True),
+        sa.Column("created_by", sa.String(length=255), nullable=True),
+        sa.Column(
+            "created_at",
+            sa.DateTime(timezone=True),
+            nullable=False,
+        ),
+        sa.Column(
+            "updated_at",
+            sa.DateTime(timezone=True),
+            nullable=False,
+        ),
+        sa.PrimaryKeyConstraint(
+            "id",
+            name=op.f("pk_connector_sanctions_acquisition_runs"),
+        ),
+    )
+    for column in (
+        "tenant_id",
+        "provider_id",
+        "source_id",
+        "status",
+        "started_at",
+        "snapshot_id",
+        "created_by",
+    ):
+        op.create_index(
+            op.f(
+                "ix_connector_sanctions_acquisition_runs_"
+                f"{column}"
+            ),
+            "connector_sanctions_acquisition_runs",
+            [column],
+        )
+    op.create_index(
+        "ix_connector_sanctions_run_health",
+        "connector_sanctions_acquisition_runs",
+        ["tenant_id", "provider_id", "status", "started_at"],
+    )
+
+    op.create_table(
+        "connector_sanctions_snapshots",
+        sa.Column("id", sa.String(length=36), nullable=False),
+        sa.Column("tenant_id", sa.String(length=36), nullable=False),
+        sa.Column("provider_id", sa.String(length=100), nullable=False),
+        sa.Column("publisher", sa.String(length=300), nullable=False),
+        sa.Column("jurisdiction", sa.String(length=100), nullable=False),
+        sa.Column("list_type", sa.String(length=100), nullable=False),
+        sa.Column("source_id", sa.String(length=200), nullable=False),
+        sa.Column(
+            "source_version",
+            sa.String(length=255),
+            nullable=False,
+        ),
+        sa.Column(
+            "publication_at",
+            sa.DateTime(timezone=True),
+            nullable=True,
+        ),
+        sa.Column(
+            "effective_at",
+            sa.DateTime(timezone=True),
+            nullable=True,
+        ),
+        sa.Column(
+            "acquired_at",
+            sa.DateTime(timezone=True),
+            nullable=False,
+        ),
+        sa.Column("source_url", sa.String(length=1500), nullable=True),
+        sa.Column(
+            "content_type",
+            sa.String(length=200),
+            nullable=False,
+        ),
+        sa.Column("byte_count", sa.Integer(), nullable=False),
+        sa.Column("sha256", sa.String(length=64), nullable=False),
+        sa.Column("signature_evidence", sa.JSON(), nullable=False),
+        sa.Column(
+            "parser_version",
+            sa.String(length=100),
+            nullable=False,
+        ),
+        sa.Column("licence_notes", sa.Text(), nullable=True),
+        sa.Column("trust_notes", sa.Text(), nullable=True),
+        sa.Column(
+            "connector_run_id",
+            sa.String(length=36),
+            nullable=False,
+        ),
+        sa.Column("transport_evidence", sa.JSON(), nullable=False),
+        sa.Column("raw_content", sa.LargeBinary(), nullable=False),
+        sa.Column(
+            "created_at",
+            sa.DateTime(timezone=True),
+            nullable=False,
+        ),
+        sa.Column(
+            "updated_at",
+            sa.DateTime(timezone=True),
+            nullable=False,
+        ),
+        sa.ForeignKeyConstraint(
+            ["connector_run_id"],
+            ["connector_sanctions_acquisition_runs.id"],
+            name=op.f(
+                "fk_connector_sanctions_snapshots_connector_run_id_"
+                "connector_sanctions_acquisition_runs"
+            ),
+            ondelete="RESTRICT",
+        ),
+        sa.PrimaryKeyConstraint(
+            "id",
+            name=op.f("pk_connector_sanctions_snapshots"),
+        ),
+        sa.UniqueConstraint(
+            "connector_run_id",
+            name="uq_connector_sanctions_snapshot_run",
+        ),
+    )
+    for column in (
+        "tenant_id",
+        "provider_id",
+        "jurisdiction",
+        "list_type",
+        "source_id",
+        "source_version",
+        "acquired_at",
+        "sha256",
+        "connector_run_id",
+    ):
+        op.create_index(
+            op.f(f"ix_connector_sanctions_snapshots_{column}"),
+            "connector_sanctions_snapshots",
+            [column],
+        )
+    op.create_index(
+        "ix_connector_sanctions_snapshot_source",
+        "connector_sanctions_snapshots",
+        ["tenant_id", "provider_id", "acquired_at"],
+    )
+    op.create_index(
+        "ix_connector_sanctions_snapshot_version",
+        "connector_sanctions_snapshots",
+        ["provider_id", "source_id", "source_version"],
+    )
+
+
+def downgrade() -> None:
+    op.drop_table("connector_sanctions_snapshots")
+    op.drop_table("connector_sanctions_acquisition_runs")

src/govoplan_connectors/backend/provider_state.py

@@ -0,0 +1,207 @@
+from __future__ import annotations
+
+from collections import defaultdict
+from datetime import UTC, datetime
+from hashlib import sha256
+
+from sqlalchemy import func, select
+from sqlalchemy.orm import Session
+
+from govoplan_connectors.backend.db.models import (
+    ConnectorSanctionsAcquisitionRun,
+    ConnectorSanctionsSnapshot,
+    ConnectorTabularSource,
+)
+from govoplan_core.core.provider_governance import (
+    ExternalProviderRuntimeState,
+    ExternalProviderStateContext,
+)
+
+
+TABULAR_PROVIDER_ID = "connectors.tabular_snapshot"
+SANCTIONS_PROVIDER_ID = "connectors.sanctions_snapshot"
+
+
+def tabular_provider_states(
+    context: ExternalProviderStateContext,
+) -> tuple[ExternalProviderRuntimeState, ...]:
+    session = _session(context)
+    statement = select(ConnectorTabularSource).where(
+        ConnectorTabularSource.deleted_at.is_(None)
+    )
+    if context.tenant_id is not None:
+        statement = statement.where(
+            ConnectorTabularSource.tenant_id == context.tenant_id
+        )
+    sources = tuple(
+        session.scalars(
+            statement.order_by(
+                ConnectorTabularSource.tenant_id,
+                ConnectorTabularSource.id,
+            ).limit(context.max_items + 1)
+        )
+    )
+    observed_at = datetime.now(UTC)
+    return tuple(_tabular_state(item, observed_at=observed_at) for item in sources)
+
+
+def sanctions_provider_states(
+    context: ExternalProviderStateContext,
+) -> tuple[ExternalProviderRuntimeState, ...]:
+    session = _session(context)
+    statement = select(ConnectorSanctionsAcquisitionRun)
+    if context.tenant_id is not None:
+        statement = statement.where(
+            ConnectorSanctionsAcquisitionRun.tenant_id == context.tenant_id
+        )
+    runs = tuple(
+        session.scalars(
+            statement.order_by(
+                ConnectorSanctionsAcquisitionRun.tenant_id,
+                ConnectorSanctionsAcquisitionRun.provider_id,
+                ConnectorSanctionsAcquisitionRun.source_id,
+                ConnectorSanctionsAcquisitionRun.started_at.desc(),
+            ).limit(max(context.max_items * 10, context.max_items + 1))
+        )
+    )
+    latest_by_binding: dict[tuple[str, str, str], ConnectorSanctionsAcquisitionRun] = {}
+    for run in runs:
+        key = (run.tenant_id, run.provider_id, run.source_id)
+        latest_by_binding.setdefault(key, run)
+        if len(latest_by_binding) >= context.max_items + 1:
+            break
+
+    snapshot_counts = _snapshot_counts(
+        session,
+        binding_keys=tuple(latest_by_binding),
+    )
+    observed_at = datetime.now(UTC)
+    return tuple(
+        _sanctions_state(
+            run,
+            observed_at=observed_at,
+            snapshot_count=snapshot_counts.get(key, 0),
+        )
+        for key, run in latest_by_binding.items()
+    )
+
+
+def _session(context: ExternalProviderStateContext) -> Session:
+    if not isinstance(context.session, Session):
+        raise RuntimeError("Connectors provider state requires a database session.")
+    return context.session
+
+
+def _tabular_state(
+    source: ConnectorTabularSource,
+    *,
+    observed_at: datetime,
+) -> ExternalProviderRuntimeState:
+    active = source.status == "active"
+    return ExternalProviderRuntimeState(
+        provider_id=TABULAR_PROVIDER_ID,
+        binding_ref=f"connectors:tabular-source:{source.id}",
+        authority_mode="external_mirror",
+        observed_at=observed_at,
+        configured=True,
+        active=active,
+        health="healthy" if active else "inactive",
+        freshness="not_applicable",
+        conflict="not_applicable",
+        recovery="ready" if active else "not_applicable",
+        last_success_at=_aware(source.updated_at or source.created_at),
+        detail=(
+            "Immutable tabular snapshot is available."
+            if active
+            else "Immutable tabular snapshot is inactive."
+        ),
+        metrics={
+            "row_count": int(source.row_count),
+            "byte_count": int(source.byte_count),
+            "schema_version": int(source.schema_version),
+        },
+    )
+
+
+def _snapshot_counts(
+    session: Session,
+    *,
+    binding_keys: tuple[tuple[str, str, str], ...],
+) -> dict[tuple[str, str, str], int]:
+    if not binding_keys:
+        return {}
+    tenant_ids = {item[0] for item in binding_keys}
+    rows = session.execute(
+        select(
+            ConnectorSanctionsSnapshot.tenant_id,
+            ConnectorSanctionsSnapshot.provider_id,
+            ConnectorSanctionsSnapshot.source_id,
+            func.count(ConnectorSanctionsSnapshot.id),
+        )
+        .where(ConnectorSanctionsSnapshot.tenant_id.in_(tenant_ids))
+        .group_by(
+            ConnectorSanctionsSnapshot.tenant_id,
+            ConnectorSanctionsSnapshot.provider_id,
+            ConnectorSanctionsSnapshot.source_id,
+        )
+    )
+    return {
+        (str(tenant_id), str(provider_id), str(source_id)): int(count)
+        for tenant_id, provider_id, source_id, count in rows
+        if (str(tenant_id), str(provider_id), str(source_id)) in binding_keys
+    }
+
+
+def _sanctions_state(
+    run: ConnectorSanctionsAcquisitionRun,
+    *,
+    observed_at: datetime,
+    snapshot_count: int,
+) -> ExternalProviderRuntimeState:
+    status = str(run.status)
+    success = status in {"succeeded", "success", "not_modified"}
+    running = status in {"running", "pending", "retry"}
+    has_snapshot = bool(run.snapshot_id) or snapshot_count > 0
+    health = "healthy" if success else "warning" if running else "error"
+    binding_digest = sha256(
+        f"{run.tenant_id}\0{run.provider_id}\0{run.source_id}".encode("utf-8")
+    ).hexdigest()[:24]
+    return ExternalProviderRuntimeState(
+        provider_id=SANCTIONS_PROVIDER_ID,
+        binding_ref=f"connectors:sanctions-source:{binding_digest}",
+        authority_mode="external_mirror",
+        observed_at=observed_at,
+        configured=True,
+        active=True,
+        health=health,
+        freshness="unknown",
+        conflict="not_applicable",
+        recovery="ready" if success and has_snapshot else "attention",
+        last_success_at=_aware(run.finished_at) if success else None,
+        detail=(
+            "Latest sanctions acquisition completed."
+            if success
+            else "Sanctions acquisition is in progress."
+            if running
+            else "Latest sanctions acquisition failed; prior accepted snapshots remain separate evidence."
+        ),
+        metrics={
+            "latest_status": status,
+            "attempt_count": int(run.attempt_count),
+            "accepted_snapshots": int(snapshot_count),
+        },
+    )
+
+
+def _aware(value: datetime | None) -> datetime | None:
+    if value is None:
+        return None
+    return value.replace(tzinfo=UTC) if value.tzinfo is None else value.astimezone(UTC)
+
+
+__all__ = [
+    "SANCTIONS_PROVIDER_ID",
+    "TABULAR_PROVIDER_ID",
+    "sanctions_provider_states",
+    "tabular_provider_states",
+]

src/govoplan_connectors/backend/recovery.py

@@ -0,0 +1,381 @@
+from __future__ import annotations
+
+from dataclasses import dataclass
+import hashlib
+from typing import Any
+from uuid import NAMESPACE_URL, uuid4, uuid5
+
+from sqlalchemy.orm import Session, sessionmaker
+
+from govoplan_core.core.recovery import (
+    RecoveryGuaranteeError,
+    RecoveryMode,
+    RecoveryPlan,
+    RecoveryStatus,
+)
+from govoplan_core.core.recovery_runtime import (
+    DurableRecoveryOperation,
+    RecoveryOperationBusy,
+    RecoveryOperationStateConflict,
+    begin_durable_recovery_operation,
+)
+from govoplan_core.core.runtime_coordination import process_runtime_identity
+
+
+class ConnectorRecoveryError(RuntimeError):
+    pass
+
+
+@dataclass(frozen=True, slots=True)
+class ConnectorRecoveryDeclaration:
+    operation_type: str
+    mode: RecoveryMode
+    provider_mutation: bool
+    idempotency: str
+    verification: tuple[str, ...]
+    recovery: tuple[str, ...]
+    implemented: bool
+
+
+CONNECTOR_RECOVERY_OPERATIONS = (
+    ConnectorRecoveryDeclaration(
+        operation_type="read-snapshot",
+        mode=RecoveryMode.ATOMIC,
+        provider_mutation=False,
+        idempotency=(
+            "Caller-supplied request keys replay a committed immutable snapshot; "
+            "otherwise each deliberate acquisition receives a generated key."
+        ),
+        verification=(
+            "provider revision or conditional cursor is recorded before fetch",
+            "domain snapshot and terminal recovery checkpoint commit together",
+            "stored bytes and provider evidence are checksum verified",
+        ),
+        recovery=(
+            "a stale running transaction is failed after its database transaction rolls back",
+            "a new deliberate acquisition may then use a new request key",
+        ),
+        implemented=True,
+    ),
+    ConnectorRecoveryDeclaration(
+        operation_type="external-mutation",
+        mode=RecoveryMode.FORWARD_RECOVERY,
+        provider_mutation=True,
+        idempotency="A stable caller key and canonical request digest are mandatory.",
+        verification=(
+            "record the remote revision and bounded provider result",
+            "verify the provider state before reporting success",
+        ),
+        recovery=(
+            "unknown outcomes remain unresolved until provider-backed reconciliation",
+            "never retry the same remote effect solely to reconstruct local state",
+        ),
+        implemented=False,
+    ),
+)
+
+
+def connector_session_factory(session: Session) -> sessionmaker[Session]:
+    bind = session.get_bind()
+    if bind is None:
+        raise ConnectorRecoveryError("Connector recovery requires a bound database session")
+    return sessionmaker(bind=bind, expire_on_commit=False)
+
+
+def _digest(value: str) -> str:
+    return hashlib.sha256(value.encode("utf-8")).hexdigest()
+
+
+def _clean_key(value: str | None) -> str:
+    clean = str(value or "").strip()
+    if clean and len(clean) > 500:
+        raise ConnectorRecoveryError("Connector idempotency keys are limited to 500 characters")
+    return clean or str(uuid4())
+
+
+def _stable_resource_id(
+    *,
+    tenant_id: str,
+    provider_id: str,
+    operation_type: str,
+    request_key: str,
+) -> str:
+    return str(
+        uuid5(
+            NAMESPACE_URL,
+            f"govoplan:{tenant_id}:{provider_id}:{operation_type}:{request_key}",
+        )
+    )
+
+
+@dataclass(slots=True)
+class ConnectorReadSnapshotRecovery:
+    operation: DurableRecoveryOperation | None
+    operation_id: str
+    request_key: str
+    resource_id: str
+    replayed: bool
+
+    def commit_success(self, session: Session, *, evidence: dict[str, Any]) -> None:
+        if self.operation is None:
+            raise ConnectorRecoveryError("A replayed connector read cannot be committed again")
+        try:
+            self.operation.commit_atomic_success(session, evidence=evidence)
+        except Exception as exc:
+            raise ConnectorRecoveryError(
+                "The connector snapshot and recovery evidence did not commit atomically"
+            ) from exc
+
+    def commit_failure(
+        self,
+        session: Session,
+        *,
+        summary: str,
+        evidence: dict[str, Any],
+    ) -> None:
+        if self.operation is None:
+            raise ConnectorRecoveryError("A replayed connector read cannot be failed again")
+        try:
+            self.operation.commit_atomic_failure(
+                session,
+                summary=summary,
+                evidence=evidence,
+            )
+        except Exception as exc:
+            raise ConnectorRecoveryError(
+                "The connector failure evidence did not commit atomically"
+            ) from exc
+
+    def fail_without_projection(
+        self,
+        *,
+        summary: str,
+        code: str,
+    ) -> None:
+        if self.operation is None:
+            return
+        self.operation.fail(
+            summary=summary,
+            evidence={
+                "verified": True,
+                "checks": {
+                    "provider_mutation": False,
+                    "projection_committed": False,
+                    "failure_code": code,
+                },
+            },
+        )
+
+
+def begin_connector_read_snapshot(
+    session: Session,
+    *,
+    tenant_id: str,
+    provider_id: str,
+    idempotency_key: str | None,
+    source_revision: str | None,
+    cursor: str | None,
+    dry_run_evidence: dict[str, Any],
+    request_metadata: dict[str, Any] | None = None,
+    resource_type: str = "connector_sync_run",
+) -> ConnectorReadSnapshotRecovery:
+    request_key = _clean_key(idempotency_key)
+    resource_id = _stable_resource_id(
+        tenant_id=tenant_id,
+        provider_id=provider_id,
+        operation_type="read-snapshot",
+        request_key=request_key,
+    )
+    request = {
+        "tenant_id": tenant_id,
+        "provider_id": provider_id,
+        "dry_run": dry_run_evidence,
+        "request_key_sha256": _digest(request_key),
+        **dict(request_metadata or {}),
+    }
+    try:
+        started = begin_durable_recovery_operation(
+            connector_session_factory(session),
+            identity=process_runtime_identity(),
+            module_id="connectors",
+            operation_type="read-snapshot",
+            idempotency_key=f"connector-read:{_digest(f'{tenant_id}:{provider_id}:{request_key}')}",
+            request=request,
+            recovery_plan=RecoveryPlan(
+                mode=RecoveryMode.ATOMIC,
+                preconditions=(
+                    "the actor is authorized for the connector source",
+                    "the provider request is read-only",
+                    "the source revision, cursor, and dry-run decision are durable",
+                ),
+                verification_steps=(
+                    "validate the bounded provider response and source revision",
+                    "commit the immutable snapshot and terminal checkpoint atomically",
+                    "compare the stored content digest with the acquired bytes",
+                ),
+            ),
+            precondition_evidence={
+                "provider_id": provider_id,
+                "source_revision": source_revision,
+                "cursor_sha256": _digest(cursor) if cursor else None,
+                "dry_run": dry_run_evidence,
+                "provider_mutation": False,
+            },
+            lease_resource_key=f"connectors:read:{tenant_id}:{_digest(provider_id)[:40]}",
+            lease_ttl_seconds=15 * 60,
+            resource_type=resource_type,
+            resource_id=resource_id,
+            metadata={
+                "resources": ["postgresql", "external-provider"],
+                "provider_mutation": False,
+                "recovery_declaration": "read-snapshot",
+            },
+        )
+    except RecoveryOperationBusy as exc:
+        raise ConnectorRecoveryError(
+            "Another runtime is already acquiring this connector source"
+        ) from exc
+    except RecoveryOperationStateConflict as exc:
+        raise ConnectorRecoveryError(
+            "This connector request is active or unresolved; reconcile it before retrying"
+        ) from exc
+    except (RecoveryGuaranteeError, RuntimeError) as exc:
+        raise ConnectorRecoveryError(
+            "The connector recovery ledger is unavailable; the provider was not contacted"
+        ) from exc
+    return ConnectorReadSnapshotRecovery(
+        operation=started.operation,
+        operation_id=started.operation_id,
+        request_key=request_key,
+        resource_id=resource_id,
+        replayed=started.replayed,
+    )
+
+
+@dataclass(slots=True)
+class ConnectorExternalMutationRecovery:
+    operation: DurableRecoveryOperation | None
+    operation_id: str
+    replayed: bool
+
+    def succeed(self, *, provider_evidence: dict[str, Any]) -> None:
+        if self.operation is not None:
+            self.operation.succeed(evidence=provider_evidence)
+
+    def reject(self, *, summary: str, provider_code: str) -> None:
+        if self.operation is not None:
+            self.operation.reject(
+                summary=summary,
+                evidence={
+                    "verified": True,
+                    "checks": {"provider_rejection": provider_code},
+                },
+            )
+
+    def outcome_unknown(self, *, summary: str, provider_code: str) -> None:
+        if self.operation is not None:
+            self.operation.unresolved(
+                status=RecoveryStatus.OUTCOME_UNKNOWN,
+                summary=summary,
+                evidence={"effect_started": True, "provider_code": provider_code},
+                failure_summary="Inspect provider state before any retry",
+            )
+
+
+def begin_connector_external_mutation(
+    session: Session,
+    *,
+    tenant_id: str,
+    provider_id: str,
+    idempotency_key: str,
+    request_sha256: str,
+    source_revision: str | None,
+    cursor: str | None,
+    dry_run_evidence: dict[str, Any],
+    resource_type: str,
+    resource_id: str,
+) -> ConnectorExternalMutationRecovery:
+    if not str(idempotency_key or "").strip():
+        raise ConnectorRecoveryError("External connector mutations require an idempotency key")
+    request_key = _clean_key(idempotency_key)
+    if len(request_sha256) != 64 or any(
+        character not in "0123456789abcdefABCDEF" for character in request_sha256
+    ):
+        raise ConnectorRecoveryError("External connector mutations require a SHA-256 request digest")
+    try:
+        started = begin_durable_recovery_operation(
+            connector_session_factory(session),
+            identity=process_runtime_identity(),
+            module_id="connectors",
+            operation_type="external-mutation",
+            idempotency_key=f"connector-write:{_digest(f'{tenant_id}:{provider_id}:{request_key}')}",
+            request={
+                "tenant_id": tenant_id,
+                "provider_id": provider_id,
+                "request_sha256": request_sha256,
+                "source_revision": source_revision,
+                "cursor_sha256": _digest(cursor) if cursor else None,
+                "dry_run": dry_run_evidence,
+            },
+            recovery_plan=RecoveryPlan(
+                mode=RecoveryMode.FORWARD_RECOVERY,
+                preconditions=(
+                    "the actor and effective connector policy authorize the mutation",
+                    "a stable idempotency key and canonical request digest are present",
+                    "the dry-run and source revision evidence are durable",
+                ),
+                forward_recovery_steps=(
+                    "inspect provider state without repeating the mutation",
+                    "record whether the provider accepted the requested revision",
+                    "retry only under a new deliberate key when absence is proven",
+                ),
+                verification_steps=(
+                    "compare provider identity and revision with the canonical request",
+                    "verify the consuming domain state independently",
+                ),
+            ),
+            precondition_evidence={
+                "request_sha256": request_sha256,
+                "source_revision": source_revision,
+                "cursor_sha256": _digest(cursor) if cursor else None,
+                "dry_run": dry_run_evidence,
+                "provider_mutation": True,
+            },
+            lease_resource_key=(
+                f"connectors:write:{tenant_id}:{_digest(provider_id)[:24]}:"
+                f"{_digest(resource_id)[:24]}"
+            ),
+            lease_ttl_seconds=15 * 60,
+            resource_type=resource_type,
+            resource_id=resource_id,
+            metadata={
+                "resources": ["postgresql", "queue", "external-provider"],
+                "provider_mutation": True,
+                "recovery_declaration": "external-mutation",
+            },
+        )
+    except (RecoveryOperationBusy, RecoveryOperationStateConflict) as exc:
+        raise ConnectorRecoveryError(
+            "This external connector effect is active or unresolved"
+        ) from exc
+    except (RecoveryGuaranteeError, RuntimeError) as exc:
+        raise ConnectorRecoveryError(
+            "The connector recovery ledger is unavailable; no external mutation started"
+        ) from exc
+    return ConnectorExternalMutationRecovery(
+        operation=started.operation,
+        operation_id=started.operation_id,
+        replayed=started.replayed,
+    )
+
+
+__all__ = [
+    "CONNECTOR_RECOVERY_OPERATIONS",
+    "ConnectorExternalMutationRecovery",
+    "ConnectorReadSnapshotRecovery",
+    "ConnectorRecoveryDeclaration",
+    "ConnectorRecoveryError",
+    "begin_connector_external_mutation",
+    "begin_connector_read_snapshot",
+    "connector_session_factory",
+]

src/govoplan_connectors/backend/router.py

@@ -0,0 +1,689 @@
+from __future__ import annotations
+
+from dataclasses import asdict
+import hashlib
+from typing import Annotated
+
+from fastapi import APIRouter, Depends, Header, HTTPException, Query, Response, status
+from sqlalchemy.orm import Session
+
+from govoplan_core.audit.logging import audit_event
+from govoplan_core.auth import ApiPrincipal, get_api_principal, has_scope
+from govoplan_core.core.tabular_sources import (
+    TabularReadRequest,
+    TabularSnapshotInput,
+    TabularSource,
+    TabularSourceAccessError,
+    TabularSourceError,
+    TabularSourceNotFoundError,
+    TabularSourceUnavailableError,
+)
+from govoplan_core.core.feeds import (
+    FeedCapabilityError,
+    FeedEntry,
+    FeedRenderRequest,
+)
+from govoplan_core.core.sanctions import SanctionsSnapshotReference
+from govoplan_core.db.session import get_session
+from govoplan_connectors.backend.schemas import (
+    FeedAcquireRequest,
+    FeedDocumentResponse,
+    FeedImportRequest,
+    FeedRenderPayload,
+    SanctionsAcquisitionRunListResponse,
+    SanctionsAcquisitionRunResponse,
+    SanctionsRefreshResponse,
+    SanctionsSnapshotListResponse,
+    SanctionsSnapshotResponse,
+    SanctionsSourceListResponse,
+    SanctionsSourceResponse,
+    SnapshotCreateRequest,
+    TabularColumnResponse,
+    TabularHealthResponse,
+    TabularPreviewDiagnosticResponse,
+    TabularPushdownResponse,
+    TabularSourceDeleteResponse,
+    TabularSourceListResponse,
+    TabularSourcePreviewResponse,
+    TabularSourceResponse,
+)
+from govoplan_connectors.backend.feeds import ConnectorFeedProvider, feed_rows
+from govoplan_connectors.backend.recovery import (
+    ConnectorRecoveryError,
+    begin_connector_read_snapshot,
+)
+from govoplan_connectors.backend.sanctions_sources import (
+    SANCTIONS_READ_SCOPE,
+    SANCTIONS_REFRESH_SCOPE,
+    SanctionsSourceAccessError,
+    SanctionsSourceError,
+    SanctionsSourceNotFoundError,
+    SqlSanctionsSnapshotProvider,
+)
+from govoplan_connectors.backend.tabular_sources import (
+    ADMIN_SCOPE,
+    READ_SCOPE,
+    WRITE_SCOPE,
+    SqlTabularSourceProvider,
+    parse_csv_snapshot,
+)
+
+
+router = APIRouter(prefix="/connectors", tags=["connectors"])
+provider = SqlTabularSourceProvider()
+sanctions_provider = SqlSanctionsSnapshotProvider()
+feed_transport = ConnectorFeedProvider()
+
+
+def _require_any_scope(principal: ApiPrincipal, *scopes: str) -> None:
+    if any(has_scope(principal, scope) for scope in scopes):
+        return
+    raise HTTPException(
+        status_code=status.HTTP_403_FORBIDDEN,
+        detail=f"Missing one of the required scopes: {', '.join(scopes)}",
+    )
+
+
+def _http_error(exc: TabularSourceError) -> HTTPException:
+    if isinstance(exc, TabularSourceNotFoundError):
+        return HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail=str(exc))
+    if isinstance(exc, TabularSourceAccessError):
+        return HTTPException(status_code=status.HTTP_403_FORBIDDEN, detail=str(exc))
+    if isinstance(exc, TabularSourceUnavailableError):
+        return HTTPException(
+            status_code=status.HTTP_503_SERVICE_UNAVAILABLE,
+            detail=str(exc),
+        )
+    return HTTPException(status_code=status.HTTP_422_UNPROCESSABLE_CONTENT, detail=str(exc))
+
+
+def _sanctions_http_error(
+    exc: SanctionsSourceError,
+) -> HTTPException:
+    if isinstance(exc, SanctionsSourceNotFoundError):
+        return HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND,
+            detail=str(exc),
+        )
+    if isinstance(exc, SanctionsSourceAccessError):
+        return HTTPException(
+            status_code=status.HTTP_403_FORBIDDEN,
+            detail=str(exc),
+        )
+    return HTTPException(
+        status_code=status.HTTP_422_UNPROCESSABLE_CONTENT,
+        detail=str(exc),
+    )
+
+
+def _feed_http_error(exc: FeedCapabilityError) -> HTTPException:
+    return HTTPException(
+        status_code=status.HTTP_422_UNPROCESSABLE_CONTENT,
+        detail=str(exc),
+    )
+
+
+def _recovery_http_error(exc: ConnectorRecoveryError) -> HTTPException:
+    detail = str(exc)
+    return HTTPException(
+        status_code=(
+            status.HTTP_409_CONFLICT
+            if "already" in detail.casefold() or "active" in detail.casefold()
+            else status.HTTP_503_SERVICE_UNAVAILABLE
+        ),
+        detail=detail,
+    )
+
+
+@router.post("/feeds/preview", response_model=FeedDocumentResponse)
+def api_preview_feed(
+    payload: FeedAcquireRequest,
+    principal: ApiPrincipal = Depends(get_api_principal),
+) -> FeedDocumentResponse:
+    _require_any_scope(principal, READ_SCOPE, ADMIN_SCOPE)
+    try:
+        document = feed_transport.fetch(
+            payload.url,
+            max_entries=payload.max_entries,
+        )
+    except FeedCapabilityError as exc:
+        raise _feed_http_error(exc) from exc
+    return FeedDocumentResponse.model_validate(asdict(document))
+
+
+@router.post(
+    "/feeds/import",
+    response_model=TabularSourceResponse,
+    status_code=status.HTTP_201_CREATED,
+)
+def api_import_feed_snapshot(
+    payload: FeedImportRequest,
+    session: Session = Depends(get_session),
+    principal: ApiPrincipal = Depends(get_api_principal),
+    idempotency_key: Annotated[
+        str | None,
+        Header(alias="Idempotency-Key", max_length=500),
+    ] = None,
+) -> TabularSourceResponse:
+    _require_any_scope(principal, WRITE_SCOPE, ADMIN_SCOPE)
+    try:
+        recovery = begin_connector_read_snapshot(
+            session,
+            tenant_id=principal.tenant_id,
+            provider_id="connectors.feed_snapshot",
+            idempotency_key=idempotency_key,
+            source_revision=None,
+            cursor=None,
+            dry_run_evidence={
+                "performed": False,
+                "reason": "read-only acquisition into an immutable snapshot",
+            },
+            request_metadata={
+                "source_url_sha256": hashlib.sha256(
+                    payload.url.encode("utf-8")
+                ).hexdigest(),
+                "source_name": payload.source_name,
+                "max_entries": payload.max_entries,
+            },
+            resource_type="connector_tabular_source",
+        )
+    except ConnectorRecoveryError as exc:
+        raise _recovery_http_error(exc) from exc
+    if recovery.replayed:
+        try:
+            source = provider.get_source(
+                session,
+                principal,
+                source_ref=f"snapshot:{recovery.resource_id}",
+            )
+        except TabularSourceError as exc:
+            raise _http_error(exc) from exc
+        return _source_response(source)
+    try:
+        document = feed_transport.fetch(
+            payload.url,
+            max_entries=payload.max_entries,
+        )
+        source = provider.create_snapshot(
+            session,
+            principal,
+            snapshot=TabularSnapshotInput(
+                name=payload.name,
+                source_name=payload.source_name,
+                description=payload.description or document.description,
+                rows=feed_rows(document),
+                metadata={
+                    "import_format": document.format,
+                    "feed": {
+                        "source_url": document.source_url,
+                        "home_url": document.home_url,
+                        "acquired_at": (
+                            document.acquired_at.isoformat()
+                            if document.acquired_at
+                            else None
+                        ),
+                        "fresh_until": (
+                            document.fresh_until.isoformat()
+                            if document.fresh_until
+                            else None
+                        ),
+                        "etag": document.etag,
+                        "last_modified": document.last_modified,
+                        "content_type": document.content_type,
+                        "sha256": document.sha256,
+                    },
+                },
+            ),
+            source_id=recovery.resource_id,
+        )
+    except (FeedCapabilityError, TabularSourceError) as exc:
+        session.rollback()
+        recovery.fail_without_projection(
+            summary="The read-only feed import failed before a snapshot committed",
+            code=exc.__class__.__name__,
+        )
+        if isinstance(exc, FeedCapabilityError):
+            raise _feed_http_error(exc) from exc
+        raise _http_error(exc) from exc
+    audit_event(
+        session,
+        tenant_id=principal.tenant_id,
+        user_id=getattr(principal.user, "id", None),
+        api_key_id=principal.api_key_id,
+        action="connectors.feed_snapshot.created",
+        object_type="connector_tabular_source",
+        object_id=source.ref,
+        details={
+            "source_url": document.source_url,
+            "format": document.format,
+            "sha256": document.sha256,
+            "row_count": source.row_count,
+        },
+    )
+    try:
+        recovery.commit_success(
+            session,
+            evidence={
+                "verified": True,
+                "checks": {
+                    "snapshot_ref": source.ref,
+                    "snapshot_fingerprint": source.fingerprint,
+                    "feed_sha256": document.sha256,
+                    "row_count": source.row_count,
+                    "provider_mutation": False,
+                },
+            },
+        )
+    except ConnectorRecoveryError as exc:
+        raise _recovery_http_error(exc) from exc
+    return _source_response(source)
+
+
+@router.post("/feeds/render")
+def api_render_feed(
+    payload: FeedRenderPayload,
+    principal: ApiPrincipal = Depends(get_api_principal),
+) -> Response:
+    _require_any_scope(principal, READ_SCOPE, ADMIN_SCOPE)
+    try:
+        rendered = feed_transport.render(
+            FeedRenderRequest(
+                format=payload.format,
+                title=payload.title,
+                feed_url=payload.feed_url,
+                home_url=payload.home_url,
+                description=payload.description,
+                language=payload.language,
+                entries=tuple(
+                    FeedEntry(**item.model_dump()) for item in payload.entries
+                ),
+                allowed_visibilities=frozenset(payload.allowed_visibilities),
+            )
+        )
+    except FeedCapabilityError as exc:
+        raise _feed_http_error(exc) from exc
+    return Response(
+        content=rendered.body,
+        media_type=rendered.content_type,
+        headers={
+            "X-GovOPlaN-Feed-Included": str(rendered.included_entries),
+            "X-GovOPlaN-Feed-Excluded": str(rendered.excluded_entries),
+        },
+    )
+
+
+@router.get("/tabular-sources", response_model=TabularSourceListResponse)
+def api_list_tabular_sources(
+    query: str = Query(default="", max_length=200),
+    limit: int = Query(default=100, ge=1, le=100),
+    session: Session = Depends(get_session),
+    principal: ApiPrincipal = Depends(get_api_principal),
+) -> TabularSourceListResponse:
+    _require_any_scope(principal, READ_SCOPE, ADMIN_SCOPE)
+    sources = provider.list_sources(
+        session,
+        principal,
+        query=query,
+        limit=limit,
+    )
+    return TabularSourceListResponse(sources=[_source_response(source) for source in sources])
+
+
+@router.post(
+    "/tabular-sources/snapshots",
+    response_model=TabularSourceResponse,
+    status_code=status.HTTP_201_CREATED,
+)
+def api_create_tabular_snapshot(
+    payload: SnapshotCreateRequest,
+    session: Session = Depends(get_session),
+    principal: ApiPrincipal = Depends(get_api_principal),
+) -> TabularSourceResponse:
+    _require_any_scope(principal, WRITE_SCOPE, ADMIN_SCOPE)
+    try:
+        rows = (
+            tuple(payload.rows or ())
+            if payload.format == "json"
+            else parse_csv_snapshot(payload.csv_text or "", delimiter=payload.delimiter)
+        )
+        source = provider.create_snapshot(
+            session,
+            principal,
+            snapshot=TabularSnapshotInput(
+                name=payload.name,
+                source_name=payload.source_name,
+                description=payload.description,
+                rows=rows,
+                metadata={"import_format": payload.format},
+            ),
+        )
+    except TabularSourceError as exc:
+        raise _http_error(exc) from exc
+    audit_event(
+        session,
+        tenant_id=principal.tenant_id,
+        user_id=getattr(principal.user, "id", None),
+        api_key_id=principal.api_key_id,
+        action="connectors.tabular_snapshot.created",
+        object_type="connector_tabular_source",
+        object_id=source.ref,
+        details={
+            "provider": source.provider,
+            "source_name": source.source_name,
+            "fingerprint": source.fingerprint,
+            "row_count": source.row_count,
+        },
+    )
+    session.commit()
+    return _source_response(source)
+
+
+@router.get(
+    "/tabular-sources/{source_id}/preview",
+    response_model=TabularSourcePreviewResponse,
+)
+def api_preview_tabular_source(
+    source_id: str,
+    limit: int = Query(default=100, ge=1, le=500),
+    offset: int = Query(default=0, ge=0),
+    max_bytes: int = Query(default=1_000_000, ge=2, le=5_000_000),
+    timeout_ms: int = Query(default=2_000, ge=1, le=10_000),
+    session: Session = Depends(get_session),
+    principal: ApiPrincipal = Depends(get_api_principal),
+) -> TabularSourcePreviewResponse:
+    _require_any_scope(principal, READ_SCOPE, ADMIN_SCOPE)
+    try:
+        result = provider.read_source(
+            session,
+            principal,
+            request=TabularReadRequest(
+                source_ref=f"snapshot:{source_id}",
+                limit=limit,
+                offset=offset,
+                max_bytes=max_bytes,
+                timeout_ms=timeout_ms,
+            ),
+        )
+    except TabularSourceError as exc:
+        raise _http_error(exc) from exc
+    return TabularSourcePreviewResponse(
+        source=_source_response(result.source),
+        rows=[dict(row) for row in result.rows],
+        total_rows=result.total_rows,
+        truncated=result.truncated,
+        returned_bytes=result.returned_bytes,
+        elapsed_ms=result.elapsed_ms,
+        effective_row_limit=result.effective_row_limit,
+        effective_byte_limit=result.effective_byte_limit,
+        effective_timeout_ms=result.effective_timeout_ms,
+        diagnostics=[
+            TabularPreviewDiagnosticResponse(
+                severity=item.severity,
+                code=item.code,
+                message=item.message,
+                details=dict(item.details),
+            )
+            for item in result.diagnostics
+        ],
+    )
+
+
+@router.delete(
+    "/tabular-sources/{source_id}",
+    response_model=TabularSourceDeleteResponse,
+)
+def api_delete_tabular_source(
+    source_id: str,
+    session: Session = Depends(get_session),
+    principal: ApiPrincipal = Depends(get_api_principal),
+) -> TabularSourceDeleteResponse:
+    _require_any_scope(principal, WRITE_SCOPE, ADMIN_SCOPE)
+    source_ref = f"snapshot:{source_id}"
+    try:
+        source = provider.delete_snapshot(
+            session,
+            principal,
+            source_ref=source_ref,
+        )
+    except TabularSourceError as exc:
+        raise _http_error(exc) from exc
+    audit_event(
+        session,
+        tenant_id=principal.tenant_id,
+        user_id=getattr(principal.user, "id", None),
+        api_key_id=principal.api_key_id,
+        action="connectors.tabular_snapshot.deleted",
+        object_type="connector_tabular_source",
+        object_id=source_ref,
+        details={"source_name": source.source_name, "fingerprint": source.fingerprint},
+    )
+    session.commit()
+    return TabularSourceDeleteResponse(deleted=True, source_ref=source_ref)
+
+
+@router.get(
+    "/sanctions/sources",
+    response_model=SanctionsSourceListResponse,
+)
+def api_list_sanctions_sources(
+    principal: ApiPrincipal = Depends(get_api_principal),
+) -> SanctionsSourceListResponse:
+    _require_any_scope(
+        principal,
+        SANCTIONS_READ_SCOPE,
+        SANCTIONS_REFRESH_SCOPE,
+        ADMIN_SCOPE,
+    )
+    return SanctionsSourceListResponse(
+        sources=[
+            SanctionsSourceResponse.model_validate(
+                source,
+                from_attributes=True,
+            )
+            for source in sanctions_provider.available_sources()
+        ]
+    )
+
+
+@router.post(
+    "/sanctions/sources/{provider_id}/refresh",
+    response_model=SanctionsRefreshResponse,
+)
+def api_refresh_sanctions_source(
+    provider_id: str,
+    session: Session = Depends(get_session),
+    principal: ApiPrincipal = Depends(get_api_principal),
+    idempotency_key: Annotated[
+        str | None,
+        Header(alias="Idempotency-Key", max_length=500),
+    ] = None,
+) -> SanctionsRefreshResponse:
+    _require_any_scope(
+        principal,
+        SANCTIONS_REFRESH_SCOPE,
+        ADMIN_SCOPE,
+    )
+    try:
+        result = sanctions_provider.refresh_source(
+            session,
+            principal,
+            provider_id=provider_id,
+            idempotency_key=idempotency_key,
+        )
+    except ConnectorRecoveryError as exc:
+        raise _recovery_http_error(exc) from exc
+    except SanctionsSourceError as exc:
+        raise _sanctions_http_error(exc) from exc
+    audit_event(
+        session,
+        tenant_id=principal.tenant_id,
+        user_id=getattr(principal.user, "id", None),
+        api_key_id=principal.api_key_id,
+        action="connectors.sanctions_source.refreshed",
+        object_type="connector_sanctions_acquisition_run",
+        object_id=result.run_id,
+        details={
+            "provider_id": provider_id,
+            "status": result.status,
+            "snapshot_ref": (
+                result.snapshot.ref
+                if result.snapshot is not None
+                else None
+            ),
+        },
+    )
+    session.commit()
+    return SanctionsRefreshResponse(
+        run_id=result.run_id,
+        provider_id=result.provider_id,
+        status=result.status,
+        snapshot=(
+            _sanctions_snapshot_response(result.snapshot)
+            if result.snapshot is not None
+            else None
+        ),
+        error=result.error,
+    )
+
+
+@router.get(
+    "/sanctions/snapshots",
+    response_model=SanctionsSnapshotListResponse,
+)
+def api_list_sanctions_snapshots(
+    limit: int = Query(default=100, ge=1, le=500),
+    session: Session = Depends(get_session),
+    principal: ApiPrincipal = Depends(get_api_principal),
+) -> SanctionsSnapshotListResponse:
+    _require_any_scope(
+        principal,
+        SANCTIONS_READ_SCOPE,
+        ADMIN_SCOPE,
+    )
+    try:
+        snapshots = sanctions_provider.list_snapshots(
+            session,
+            principal,
+            limit=limit,
+        )
+    except SanctionsSourceError as exc:
+        raise _sanctions_http_error(exc) from exc
+    return SanctionsSnapshotListResponse(
+        snapshots=[
+            _sanctions_snapshot_response(item)
+            for item in snapshots
+        ]
+    )
+
+
+@router.get(
+    "/sanctions/runs",
+    response_model=SanctionsAcquisitionRunListResponse,
+)
+def api_list_sanctions_runs(
+    limit: int = Query(default=100, ge=1, le=500),
+    session: Session = Depends(get_session),
+    principal: ApiPrincipal = Depends(get_api_principal),
+) -> SanctionsAcquisitionRunListResponse:
+    _require_any_scope(
+        principal,
+        SANCTIONS_READ_SCOPE,
+        ADMIN_SCOPE,
+    )
+    try:
+        runs = sanctions_provider.list_runs(
+            session,
+            principal,
+            limit=limit,
+        )
+    except SanctionsSourceError as exc:
+        raise _sanctions_http_error(exc) from exc
+    return SanctionsAcquisitionRunListResponse(
+        runs=[
+            SanctionsAcquisitionRunResponse.model_validate(
+                item,
+                from_attributes=True,
+            )
+            for item in runs
+        ]
+    )
+
+
+def _source_response(source: TabularSource) -> TabularSourceResponse:
+    return TabularSourceResponse(
+        ref=source.ref,
+        provider=source.provider,
+        source_name=source.source_name,
+        name=source.name,
+        description=source.description,
+        columns=[
+            TabularColumnResponse(
+                name=column.name,
+                data_type=column.data_type,
+                nullable=column.nullable,
+            )
+            for column in source.schema
+        ],
+        schema_version=source.schema_version,
+        fingerprint=source.fingerprint,
+        row_count=source.row_count,
+        byte_count=source.byte_count,
+        updated_at=source.updated_at.isoformat() if source.updated_at else None,
+        capabilities=list(source.capabilities),
+        metadata=dict(source.metadata),
+        source_mode=source.source_mode,
+        pushdown=TabularPushdownResponse(
+            projections=source.pushdown.projections,
+            pagination=source.pushdown.pagination,
+            filters=list(source.pushdown.filters),
+            aggregations=list(source.pushdown.aggregations),
+            sorting=list(source.pushdown.sorting),
+        ),
+        health=TabularHealthResponse(
+            status=source.health.status,
+            code=source.health.code,
+            summary=source.health.summary,
+            checked_at=(
+                source.health.checked_at.isoformat()
+                if source.health.checked_at
+                else None
+            ),
+            details=dict(source.health.details),
+        ),
+    )
+
+
+def _sanctions_snapshot_response(
+    snapshot: SanctionsSnapshotReference,
+) -> SanctionsSnapshotResponse:
+    return SanctionsSnapshotResponse.model_validate(
+        {
+            "ref": snapshot.ref,
+            "provider_id": snapshot.provider_id,
+            "publisher": snapshot.publisher,
+            "jurisdiction": snapshot.jurisdiction,
+            "list_type": snapshot.list_type,
+            "source_id": snapshot.source_id,
+            "source_version": snapshot.source_version,
+            "publication_at": snapshot.publication_at,
+            "effective_at": snapshot.effective_at,
+            "acquired_at": snapshot.acquired_at,
+            "content_type": snapshot.content_type,
+            "byte_count": snapshot.byte_count,
+            "sha256": snapshot.sha256,
+            "parser_version": snapshot.parser_version,
+            "raw_evidence_ref": snapshot.raw_evidence_ref,
+            "connector_run_id": snapshot.connector_run_id,
+            "signature_evidence": dict(
+                snapshot.signature_evidence
+            ),
+            "licence_notes": snapshot.licence_notes,
+            "trust_notes": snapshot.trust_notes,
+            "transport_evidence": dict(
+                snapshot.transport_evidence
+            ),
+        }
+    )
+
+
+__all__ = ["router"]

src/govoplan_connectors/backend/schemas.py

@@ -0,0 +1,258 @@
+from __future__ import annotations
+
+from datetime import datetime
+from typing import Any, Literal
+
+from pydantic import BaseModel, Field, model_validator
+
+
+class FeedAcquireRequest(BaseModel):
+    url: str = Field(min_length=1, max_length=2000)
+    max_entries: int = Field(default=2_000, ge=1, le=10_000)
+
+
+class FeedImportRequest(FeedAcquireRequest):
+    name: str = Field(min_length=1, max_length=300)
+    source_name: str = Field(
+        min_length=1,
+        max_length=120,
+        pattern=r"^[A-Za-z_][A-Za-z0-9_]*$",
+    )
+    description: str | None = Field(default=None, max_length=4000)
+
+
+class FeedEntryPayload(BaseModel):
+    id: str = Field(min_length=1, max_length=2000)
+    title: str = Field(min_length=1, max_length=1000)
+    url: str | None = Field(default=None, max_length=2000)
+    summary: str | None = None
+    content: str | None = None
+    author: str | None = Field(default=None, max_length=500)
+    published_at: datetime | None = None
+    updated_at: datetime | None = None
+    categories: list[str] = Field(default_factory=list, max_length=100)
+    enclosures: list[dict[str, Any]] = Field(default_factory=list, max_length=100)
+    visibility: Literal["public", "tenant", "private"] = "public"
+    metadata: dict[str, Any] = Field(default_factory=dict)
+
+
+class FeedDocumentResponse(BaseModel):
+    format: Literal["rss", "atom"]
+    title: str
+    source_url: str
+    description: str | None = None
+    home_url: str | None = None
+    language: str | None = None
+    updated_at: datetime | None = None
+    acquired_at: datetime | None = None
+    fresh_until: datetime | None = None
+    etag: str | None = None
+    last_modified: str | None = None
+    content_type: str | None = None
+    sha256: str
+    entries: list[FeedEntryPayload]
+    metadata: dict[str, Any] = Field(default_factory=dict)
+
+
+class FeedRenderPayload(BaseModel):
+    format: Literal["rss", "atom"]
+    title: str = Field(min_length=1, max_length=1000)
+    feed_url: str = Field(min_length=1, max_length=2000)
+    home_url: str = Field(min_length=1, max_length=2000)
+    description: str | None = None
+    language: str | None = Field(default=None, max_length=100)
+    entries: list[FeedEntryPayload] = Field(default_factory=list, max_length=10_000)
+    allowed_visibilities: list[Literal["public", "tenant", "private"]] = Field(
+        default_factory=lambda: ["public"],
+        max_length=3,
+    )
+
+
+class SnapshotCreateRequest(BaseModel):
+    name: str = Field(min_length=1, max_length=300)
+    source_name: str = Field(
+        min_length=1,
+        max_length=120,
+        pattern=r"^[A-Za-z_][A-Za-z0-9_]*$",
+    )
+    description: str | None = Field(default=None, max_length=4000)
+    format: Literal["json", "csv"] = "json"
+    rows: list[dict[str, Any]] | None = Field(default=None, max_length=10_000)
+    csv_text: str | None = Field(default=None, max_length=5_000_000)
+    delimiter: Literal[",", ";", "\t", "|"] = ","
+
+    @model_validator(mode="after")
+    def validate_payload(self) -> "SnapshotCreateRequest":
+        if self.format == "json" and self.rows is None:
+            raise ValueError("JSON snapshots require rows.")
+        if self.format == "json" and self.csv_text is not None:
+            raise ValueError("JSON snapshots cannot include CSV text.")
+        if self.format == "csv" and not self.csv_text:
+            raise ValueError("CSV snapshots require CSV text.")
+        if self.format == "csv" and self.rows is not None:
+            raise ValueError("CSV snapshots cannot include JSON rows.")
+        return self
+
+
+class TabularColumnResponse(BaseModel):
+    name: str
+    data_type: str
+    nullable: bool
+
+
+class TabularPushdownResponse(BaseModel):
+    projections: bool
+    pagination: bool
+    filters: list[str]
+    aggregations: list[str]
+    sorting: list[str]
+
+
+class TabularHealthResponse(BaseModel):
+    status: Literal["healthy", "warning", "error", "unknown"]
+    code: str
+    summary: str
+    checked_at: str | None
+    details: dict[str, Any]
+
+
+class TabularPreviewDiagnosticResponse(BaseModel):
+    severity: Literal["info", "warning", "error"]
+    code: str
+    message: str
+    details: dict[str, Any]
+
+
+class TabularSourceResponse(BaseModel):
+    ref: str
+    provider: str
+    source_name: str
+    name: str
+    description: str | None
+    columns: list[TabularColumnResponse]
+    schema_version: str
+    fingerprint: str
+    row_count: int | None
+    byte_count: int | None
+    updated_at: str | None
+    capabilities: list[str]
+    metadata: dict[str, Any]
+    source_mode: Literal["live", "cached", "file_backed", "static"]
+    pushdown: TabularPushdownResponse
+    health: TabularHealthResponse
+
+
+class TabularSourceListResponse(BaseModel):
+    sources: list[TabularSourceResponse]
+
+
+class TabularSourcePreviewResponse(BaseModel):
+    source: TabularSourceResponse
+    rows: list[dict[str, Any]]
+    total_rows: int
+    truncated: bool
+    returned_bytes: int
+    elapsed_ms: int
+    effective_row_limit: int
+    effective_byte_limit: int
+    effective_timeout_ms: int
+    diagnostics: list[TabularPreviewDiagnosticResponse]
+
+
+class TabularSourceDeleteResponse(BaseModel):
+    deleted: bool
+    source_ref: str
+
+
+class SanctionsSourceResponse(BaseModel):
+    provider_id: str
+    publisher: str
+    jurisdiction: str
+    list_type: str
+    source_id: str
+    source_url: str | None
+    parser_version: str
+    licence_notes: str
+    trust_notes: str
+
+
+class SanctionsSourceListResponse(BaseModel):
+    sources: list[SanctionsSourceResponse]
+
+
+class SanctionsSnapshotResponse(BaseModel):
+    ref: str
+    provider_id: str
+    publisher: str
+    jurisdiction: str
+    list_type: str
+    source_id: str
+    source_version: str
+    publication_at: datetime | None
+    effective_at: datetime | None
+    acquired_at: datetime
+    content_type: str
+    byte_count: int
+    sha256: str
+    parser_version: str
+    raw_evidence_ref: str
+    connector_run_id: str
+    signature_evidence: dict[str, Any]
+    licence_notes: str | None
+    trust_notes: str | None
+    transport_evidence: dict[str, Any]
+
+
+class SanctionsSnapshotListResponse(BaseModel):
+    snapshots: list[SanctionsSnapshotResponse]
+
+
+class SanctionsAcquisitionRunResponse(BaseModel):
+    id: str
+    provider_id: str
+    source_id: str
+    status: str
+    attempt_count: int
+    request_evidence: dict[str, Any]
+    response_evidence: dict[str, Any]
+    started_at: datetime
+    finished_at: datetime | None
+    snapshot_id: str | None
+    error: str | None
+
+
+class SanctionsAcquisitionRunListResponse(BaseModel):
+    runs: list[SanctionsAcquisitionRunResponse]
+
+
+class SanctionsRefreshResponse(BaseModel):
+    run_id: str
+    provider_id: str
+    status: str
+    snapshot: SanctionsSnapshotResponse | None
+    error: str | None
+
+
+__all__ = [
+    "FeedAcquireRequest",
+    "FeedDocumentResponse",
+    "FeedEntryPayload",
+    "FeedImportRequest",
+    "FeedRenderPayload",
+    "SnapshotCreateRequest",
+    "SanctionsAcquisitionRunListResponse",
+    "SanctionsAcquisitionRunResponse",
+    "SanctionsRefreshResponse",
+    "SanctionsSnapshotListResponse",
+    "SanctionsSnapshotResponse",
+    "SanctionsSourceListResponse",
+    "SanctionsSourceResponse",
+    "TabularColumnResponse",
+    "TabularHealthResponse",
+    "TabularPreviewDiagnosticResponse",
+    "TabularPushdownResponse",
+    "TabularSourceDeleteResponse",
+    "TabularSourceListResponse",
+    "TabularSourcePreviewResponse",
+    "TabularSourceResponse",
+]

src/govoplan_connectors/backend/tabular_sources.py

@@ -0,0 +1,498 @@
+from __future__ import annotations
+
+import hashlib
+import json
+import time
+from collections.abc import Mapping, Sequence
+from datetime import datetime
+from decimal import Decimal
+from typing import Any, Callable
+
+from sqlalchemy import or_, select
+from sqlalchemy.orm import Session
+
+from govoplan_core.auth import ApiPrincipal, has_scope
+from govoplan_core.core.tabular_sources import (
+    TabularColumn,
+    TabularPreviewDiagnostic,
+    TabularPushdown,
+    TabularReadRequest,
+    TabularReadResult,
+    TabularSnapshotInput,
+    TabularSource,
+    TabularSourceAccessError,
+    TabularSourceNotFoundError,
+    TabularSourceHealth,
+    TabularSourceUnavailableError,
+    TabularSourceValidationError,
+    parse_tabular_csv,
+)
+from govoplan_core.db.base import utcnow
+from govoplan_connectors.backend.db.models import ConnectorTabularSource
+
+
+READ_SCOPE = "connectors:source:read"
+WRITE_SCOPE = "connectors:source:write"
+ADMIN_SCOPE = "connectors:source:admin"
+MAX_SNAPSHOT_ROWS = 10_000
+MAX_SNAPSHOT_BYTES = 5_000_000
+MAX_READ_ROWS = 500
+MAX_READ_BYTES = 1_000_000
+MAX_READ_TIMEOUT_MS = 2_000
+
+
+class SqlTabularSourceProvider:
+    def __init__(self, *, clock: Callable[[], float] = time.monotonic) -> None:
+        self._clock = clock
+
+    def list_sources(
+        self,
+        session: object,
+        principal: object,
+        *,
+        query: str = "",
+        limit: int = 100,
+    ) -> Sequence[TabularSource]:
+        db, api_principal = _context(session, principal, READ_SCOPE)
+        normalized_query = str(query or "").strip()
+        statement = select(ConnectorTabularSource).where(
+            ConnectorTabularSource.tenant_id == api_principal.tenant_id,
+            ConnectorTabularSource.deleted_at.is_(None),
+            ConnectorTabularSource.status == "active",
+        )
+        if normalized_query:
+            pattern = f"%{_escape_like(normalized_query)}%"
+            statement = statement.where(
+                or_(
+                    ConnectorTabularSource.name.ilike(pattern, escape="\\"),
+                    ConnectorTabularSource.source_name.ilike(pattern, escape="\\"),
+                )
+            )
+        statement = statement.order_by(
+            ConnectorTabularSource.updated_at.desc(),
+            ConnectorTabularSource.name,
+        ).limit(max(1, min(int(limit), 100)))
+        return tuple(_source_dto(item) for item in db.scalars(statement))
+
+    def get_source(
+        self,
+        session: object,
+        principal: object,
+        *,
+        source_ref: str,
+    ) -> TabularSource | None:
+        db, api_principal = _context(session, principal, READ_SCOPE)
+        item = _source_record(
+            db,
+            tenant_id=api_principal.tenant_id,
+            source_ref=source_ref,
+        )
+        return _source_dto(item) if item is not None else None
+
+    def read_source(
+        self,
+        session: object,
+        principal: object,
+        *,
+        request: TabularReadRequest,
+    ) -> TabularReadResult:
+        db, api_principal = _context(session, principal, READ_SCOPE)
+        item = _source_record(
+            db,
+            tenant_id=api_principal.tenant_id,
+            source_ref=request.source_ref,
+        )
+        if item is None:
+            raise TabularSourceNotFoundError("Tabular source not found.")
+        if request.expected_fingerprint and request.expected_fingerprint != item.fingerprint:
+            raise TabularSourceValidationError(
+                "The source fingerprint changed; refresh the source node before running it."
+            )
+
+        started = self._clock()
+        limit = max(1, min(int(request.limit), MAX_READ_ROWS))
+        byte_limit = max(2, min(int(request.max_bytes), MAX_READ_BYTES))
+        timeout_ms = max(1, min(int(request.timeout_ms), MAX_READ_TIMEOUT_MS))
+        offset = max(0, int(request.offset))
+        diagnostics: list[TabularPreviewDiagnostic] = []
+        if limit != request.limit:
+            diagnostics.append(
+                _preview_diagnostic(
+                    "preview.row_limit_tightened",
+                    "The provider tightened the requested row limit.",
+                    requested=request.limit,
+                    effective=limit,
+                )
+            )
+        if byte_limit != request.max_bytes:
+            diagnostics.append(
+                _preview_diagnostic(
+                    "preview.byte_limit_tightened",
+                    "The provider tightened the requested byte limit.",
+                    requested=request.max_bytes,
+                    effective=byte_limit,
+                )
+            )
+        if timeout_ms != request.timeout_ms:
+            diagnostics.append(
+                _preview_diagnostic(
+                    "preview.timeout_tightened",
+                    "The provider tightened the requested time limit.",
+                    requested=request.timeout_ms,
+                    effective=timeout_ms,
+                )
+            )
+        selected_columns = tuple(dict.fromkeys(request.columns))
+        known_columns = {column["name"] for column in item.schema_}
+        unknown_columns = [column for column in selected_columns if column not in known_columns]
+        if unknown_columns:
+            raise TabularSourceValidationError(
+                f"Unknown source columns: {', '.join(unknown_columns)}"
+            )
+        rows: list[dict[str, object]] = []
+        returned_bytes = 2
+        stopped_for = ""
+        for row in item.rows[offset:]:
+            if len(rows) >= limit:
+                stopped_for = "rows"
+                break
+            elapsed_ms = int(max(0.0, self._clock() - started) * 1_000)
+            if elapsed_ms >= timeout_ms:
+                if not rows:
+                    raise TabularSourceUnavailableError(
+                        "Tabular source preview exceeded its time budget."
+                    )
+                stopped_for = "time"
+                break
+            selected = {
+                key: value
+                for key, value in row.items()
+                if not selected_columns or key in selected_columns
+            }
+            row_bytes = len(
+                json.dumps(
+                    selected,
+                    sort_keys=True,
+                    separators=(",", ":"),
+                    default=str,
+                ).encode("utf-8")
+            )
+            additional_bytes = row_bytes + (1 if rows else 0)
+            if returned_bytes + additional_bytes > byte_limit:
+                if not rows:
+                    raise TabularSourceValidationError(
+                        "A single source row exceeds the preview byte limit."
+                    )
+                stopped_for = "bytes"
+                break
+            rows.append(selected)
+            returned_bytes += additional_bytes
+        elapsed_ms = int(max(0.0, self._clock() - started) * 1_000)
+        if stopped_for:
+            labels = {
+                "rows": ("preview.row_limit_reached", "row"),
+                "bytes": ("preview.byte_limit_reached", "byte"),
+                "time": ("preview.timeout_reached", "time"),
+            }
+            code, label = labels[stopped_for]
+            diagnostics.append(
+                TabularPreviewDiagnostic(
+                    severity="warning",
+                    code=code,
+                    message=f"The preview stopped at its effective {label} limit.",
+                )
+            )
+        return TabularReadResult(
+            source=_source_dto(item),
+            rows=tuple(rows),
+            total_rows=item.row_count,
+            truncated=offset + len(rows) < item.row_count,
+            returned_bytes=returned_bytes,
+            elapsed_ms=elapsed_ms,
+            effective_row_limit=limit,
+            effective_byte_limit=byte_limit,
+            effective_timeout_ms=timeout_ms,
+            diagnostics=tuple(diagnostics),
+        )
+
+    def create_snapshot(
+        self,
+        session: object,
+        principal: object,
+        *,
+        snapshot: TabularSnapshotInput,
+        source_id: str | None = None,
+    ) -> TabularSource:
+        db, api_principal = _context(session, principal, WRITE_SCOPE)
+        name = snapshot.name.strip()
+        source_name = snapshot.source_name.strip()
+        if not name:
+            raise TabularSourceValidationError("Snapshot name is required.")
+        if not source_name:
+            raise TabularSourceValidationError("Snapshot source name is required.")
+        if len(snapshot.rows) > MAX_SNAPSHOT_ROWS:
+            raise TabularSourceValidationError(
+                f"Snapshots are limited to {MAX_SNAPSHOT_ROWS:,} rows."
+            )
+        rows = [_json_row(row) for row in snapshot.rows]
+        encoded = json.dumps(rows, sort_keys=True, separators=(",", ":"), default=str).encode("utf-8")
+        if len(encoded) > MAX_SNAPSHOT_BYTES:
+            raise TabularSourceValidationError(
+                f"Snapshots are limited to {MAX_SNAPSHOT_BYTES // 1_000_000} MB."
+            )
+        existing = db.scalar(
+            select(ConnectorTabularSource.id).where(
+                ConnectorTabularSource.tenant_id == api_principal.tenant_id,
+                ConnectorTabularSource.source_name == source_name,
+            )
+        )
+        if existing is not None:
+            raise TabularSourceValidationError(
+                f"A tabular source named {source_name!r} already exists."
+            )
+        schema = infer_schema(rows)
+        fingerprint = snapshot_fingerprint(rows, schema)
+        actor_id = _actor_id(api_principal)
+        item = ConnectorTabularSource(
+            tenant_id=api_principal.tenant_id,
+            provider="snapshot",
+            source_name=source_name,
+            name=name,
+            description=_clean_optional(snapshot.description),
+            status="active",
+            schema_version=1,
+            schema_=[_column_payload(column) for column in schema],
+            rows=rows,
+            fingerprint=fingerprint,
+            row_count=len(rows),
+            byte_count=len(encoded),
+            metadata_=dict(snapshot.metadata),
+            created_by=actor_id,
+            updated_by=actor_id,
+        )
+        if source_id:
+            item.id = source_id
+        db.add(item)
+        db.flush()
+        return _source_dto(item)
+
+    def delete_snapshot(
+        self,
+        session: object,
+        principal: object,
+        *,
+        source_ref: str,
+    ) -> TabularSource:
+        db, api_principal = _context(session, principal, WRITE_SCOPE)
+        item = _source_record(
+            db,
+            tenant_id=api_principal.tenant_id,
+            source_ref=source_ref,
+        )
+        if item is None:
+            raise TabularSourceNotFoundError("Tabular source not found.")
+        item.deleted_at = utcnow()
+        item.status = "retired"
+        item.updated_by = _actor_id(api_principal)
+        db.flush()
+        return _source_dto(item)
+
+
+def parse_csv_snapshot(csv_text: str, *, delimiter: str) -> tuple[Mapping[str, object], ...]:
+    return parse_tabular_csv(
+        csv_text,
+        delimiter=delimiter,
+        max_rows=MAX_SNAPSHOT_ROWS,
+    )
+
+
+def infer_schema(rows: Sequence[Mapping[str, object]]) -> tuple[TabularColumn, ...]:
+    names: list[str] = []
+    for row in rows:
+        for name in row:
+            if name not in names:
+                names.append(name)
+    result: list[TabularColumn] = []
+    for name in names:
+        values = [row.get(name) for row in rows]
+        concrete = [value for value in values if value is not None]
+        data_type = _type_name(concrete[0]) if concrete else "unknown"
+        if any(_type_name(value) != data_type for value in concrete[1:]):
+            data_type = "mixed"
+        result.append(
+            TabularColumn(
+                name=name,
+                data_type=data_type,
+                nullable=len(concrete) != len(values),
+            )
+        )
+    return tuple(result)
+
+
+def snapshot_fingerprint(
+    rows: Sequence[Mapping[str, object]],
+    schema: Sequence[TabularColumn],
+) -> str:
+    payload = {
+        "schema": [_column_payload(column) for column in schema],
+        "rows": [dict(row) for row in rows],
+    }
+    encoded = json.dumps(payload, sort_keys=True, separators=(",", ":"), default=str)
+    return hashlib.sha256(encoded.encode("utf-8")).hexdigest()
+
+
+def _source_record(
+    session: Session,
+    *,
+    tenant_id: str,
+    source_ref: str,
+) -> ConnectorTabularSource | None:
+    source_id = source_ref.removeprefix("snapshot:")
+    if not source_id or source_id == source_ref:
+        return None
+    return session.scalar(
+        select(ConnectorTabularSource).where(
+            ConnectorTabularSource.id == source_id,
+            ConnectorTabularSource.tenant_id == tenant_id,
+            ConnectorTabularSource.deleted_at.is_(None),
+        )
+    )
+
+
+def _source_dto(item: ConnectorTabularSource) -> TabularSource:
+    return TabularSource(
+        ref=f"snapshot:{item.id}",
+        provider=item.provider,
+        source_name=item.source_name,
+        name=item.name,
+        description=item.description,
+        schema=tuple(TabularColumn(**column) for column in item.schema_),
+        schema_version=str(item.schema_version),
+        fingerprint=item.fingerprint,
+        row_count=item.row_count,
+        byte_count=item.byte_count,
+        updated_at=item.updated_at,
+        capabilities=("read", "preview"),
+        metadata=dict(item.metadata_),
+        source_mode="cached",
+        pushdown=TabularPushdown(
+            projections=True,
+            pagination=True,
+        ),
+        health=TabularSourceHealth(
+            status="healthy",
+            code="snapshot.ready",
+            summary="The immutable connector snapshot is ready.",
+            checked_at=item.updated_at,
+            details={"immutable": True},
+        ),
+    )
+
+
+def _preview_diagnostic(
+    code: str,
+    message: str,
+    *,
+    requested: int,
+    effective: int,
+) -> TabularPreviewDiagnostic:
+    return TabularPreviewDiagnostic(
+        severity="info",
+        code=code,
+        message=message,
+        details={"requested": requested, "effective": effective},
+    )
+
+
+def _context(
+    session: object,
+    principal: object,
+    required_scope: str,
+) -> tuple[Session, ApiPrincipal]:
+    if not isinstance(session, Session):
+        raise TypeError("Tabular source providers require a SQLAlchemy session.")
+    if not isinstance(principal, ApiPrincipal):
+        raise TabularSourceAccessError("A tenant API principal is required.")
+    accepted_scopes = {required_scope, ADMIN_SCOPE}
+    if required_scope == READ_SCOPE:
+        accepted_scopes.update(
+            {
+                WRITE_SCOPE,
+                "datasources:catalogue:read",
+                "datasources:source:write",
+                "datasources:source:admin",
+            }
+        )
+    if not any(has_scope(principal, scope) for scope in accepted_scopes):
+        raise TabularSourceAccessError(f"Missing scope: {required_scope}")
+    return session, principal
+
+
+def _json_row(row: Mapping[str, object]) -> dict[str, Any]:
+    normalized = {str(key).strip(): value for key, value in row.items()}
+    if not normalized or any(not key for key in normalized):
+        raise TabularSourceValidationError("Every snapshot row needs named columns.")
+    try:
+        json.dumps(normalized, default=_unsupported_json)
+    except (TypeError, ValueError) as exc:
+        raise TabularSourceValidationError(f"Snapshot values must be JSON compatible: {exc}") from exc
+    return json.loads(json.dumps(normalized, default=_unsupported_json))
+
+
+def _unsupported_json(value: object) -> object:
+    if isinstance(value, (datetime, Decimal)):
+        return str(value)
+    raise TypeError(f"{type(value).__name__} is not JSON serializable")
+
+
+def _type_name(value: object) -> str:
+    if isinstance(value, bool):
+        return "boolean"
+    if isinstance(value, int):
+        return "integer"
+    if isinstance(value, (float, Decimal)):
+        return "number"
+    if isinstance(value, str):
+        return "string"
+    if isinstance(value, list):
+        return "array"
+    if isinstance(value, dict):
+        return "object"
+    return type(value).__name__.lower()
+
+
+def _column_payload(column: TabularColumn) -> dict[str, object]:
+    return {
+        "name": column.name,
+        "data_type": column.data_type,
+        "nullable": column.nullable,
+    }
+
+
+def _actor_id(principal: ApiPrincipal) -> str | None:
+    return principal.account_id or principal.membership_id or principal.identity_id
+
+
+def _clean_optional(value: str | None) -> str | None:
+    cleaned = str(value or "").strip()
+    return cleaned or None
+
+
+def _escape_like(value: str) -> str:
+    return value.replace("\\", "\\\\").replace("%", "\\%").replace("_", "\\_")
+
+
+__all__ = [
+    "ADMIN_SCOPE",
+    "MAX_READ_ROWS",
+    "MAX_READ_BYTES",
+    "MAX_READ_TIMEOUT_MS",
+    "MAX_SNAPSHOT_BYTES",
+    "MAX_SNAPSHOT_ROWS",
+    "READ_SCOPE",
+    "SqlTabularSourceProvider",
+    "WRITE_SCOPE",
+    "infer_schema",
+    "parse_csv_snapshot",
+    "snapshot_fingerprint",
+]

src/govoplan_connectors/py.typed

@@ -0,0 +1 @@
+

tests/test_datasource_origins.py

@@ -0,0 +1,92 @@
+from __future__ import annotations
+
+import unittest
+
+from sqlalchemy import create_engine
+from sqlalchemy.orm import sessionmaker
+
+from govoplan_core.auth import ApiPrincipal
+from govoplan_core.core.access import PrincipalRef
+from govoplan_core.core.datasources import DatasourceOriginReadRequest
+from govoplan_core.core.tabular_sources import TabularSnapshotInput
+from govoplan_core.db.base import Base
+from govoplan_connectors.backend.datasource_origins import (
+    ConnectorDatasourceOriginProvider,
+)
+from govoplan_connectors.backend.db.models import ConnectorTabularSource
+from govoplan_connectors.backend.tabular_sources import (
+    WRITE_SCOPE,
+    SqlTabularSourceProvider,
+)
+
+
+def principal(*, scopes: tuple[str, ...]) -> ApiPrincipal:
+    return ApiPrincipal(
+        principal=PrincipalRef(
+            account_id="account-1",
+            membership_id="membership-1",
+            tenant_id="tenant-1",
+            scopes=frozenset(scopes),
+        ),
+        account=object(),
+        user=object(),
+    )
+
+
+class ConnectorDatasourceOriginTests(unittest.TestCase):
+    def setUp(self) -> None:
+        self.engine = create_engine("sqlite:///:memory:")
+        Base.metadata.create_all(
+            self.engine,
+            tables=[ConnectorTabularSource.__table__],
+        )
+        self.Session = sessionmaker(bind=self.engine)
+        self.session = self.Session()
+        provider = SqlTabularSourceProvider()
+        self.source = provider.create_snapshot(
+            self.session,
+            principal(scopes=(WRITE_SCOPE,)),
+            snapshot=TabularSnapshotInput(
+                name="Imported cases",
+                source_name="imported_cases",
+                rows=({"id": 1, "name": "Ada"},),
+            ),
+        )
+        self.session.commit()
+        self.origins = ConnectorDatasourceOriginProvider(provider)
+
+    def tearDown(self) -> None:
+        self.session.close()
+        Base.metadata.drop_all(
+            self.engine,
+            tables=[ConnectorTabularSource.__table__],
+        )
+        self.engine.dispose()
+
+    def test_datasource_reader_can_discover_and_read_connector_origin(self) -> None:
+        datasource_principal = principal(
+            scopes=("datasources:catalogue:read",),
+        )
+
+        origins = self.origins.list_origins(
+            self.session,
+            datasource_principal,
+        )
+        result = self.origins.read_origin(
+            self.session,
+            datasource_principal,
+            request=DatasourceOriginReadRequest(origin_ref=self.source.ref),
+        )
+
+        self.assertEqual((self.source.ref,), tuple(item.ref for item in origins))
+        self.assertEqual(("live", "cached"), origins[0].supported_modes)
+        self.assertEqual(({"id": 1, "name": "Ada"},), result.rows)
+        self.assertEqual("cached", origins[0].source_mode)
+        self.assertTrue(origins[0].pushdown.projections)
+        self.assertEqual("healthy", origins[0].health.status)
+        self.assertGreater(result.returned_bytes, 2)
+        self.assertEqual(1_000_000, result.effective_byte_limit)
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_feeds.py

@@ -0,0 +1,103 @@
+from __future__ import annotations
+
+import unittest
+from unittest.mock import patch
+
+from defusedxml import ElementTree as SafeET
+
+from govoplan_connectors.backend.feeds import ConnectorFeedProvider, feed_rows
+from govoplan_core.core.feeds import (
+    FeedCapabilityError,
+    FeedEntry,
+    FeedRenderRequest,
+)
+from govoplan_core.security.http_fetch import HttpFetchResponse
+
+
+RSS = b"""<?xml version="1.0"?>
+<rss version="2.0"><channel>
+  <title>Decisions</title><link>https://example.test/</link>
+  <description>Published decisions</description>
+  <item><guid>decision-1</guid><title>Decision one</title>
+    <link>https://example.test/1</link>
+    <pubDate>Fri, 31 Jul 2026 10:00:00 GMT</pubDate>
+    <category>planning</category>
+  </item>
+</channel></rss>"""
+
+ATOM = b"""<?xml version="1.0"?>
+<feed xmlns="http://www.w3.org/2005/Atom">
+  <id>https://example.test/feed</id><title>Updates</title>
+  <updated>2026-07-31T10:00:00Z</updated>
+  <link href="https://example.test/" />
+  <entry><id>update-1</id><title>Update one</title>
+    <updated>2026-07-31T10:00:00Z</updated>
+    <link href="https://example.test/update-1" />
+  </entry>
+</feed>"""
+
+
+class ConnectorFeedProviderTests(unittest.TestCase):
+    def setUp(self) -> None:
+        self.provider = ConnectorFeedProvider()
+
+    def test_rss_and_atom_are_normalized_to_tabular_entries(self) -> None:
+        rss = self.provider.parse(RSS, source_url="https://example.test/rss")
+        atom = self.provider.parse(ATOM, source_url="https://example.test/atom")
+
+        self.assertEqual("rss", rss.format)
+        self.assertEqual("decision-1", rss.entries[0].id)
+        self.assertEqual("atom", atom.format)
+        self.assertEqual("https://example.test/update-1", atom.entries[0].url)
+        self.assertEqual("decision-1", feed_rows(rss)[0]["id"])
+        self.assertEqual(64, len(rss.sha256))
+
+    def test_fetch_records_transport_freshness_and_provenance(self) -> None:
+        with patch(
+            "govoplan_connectors.backend.feeds.fetch_http",
+            return_value=HttpFetchResponse(
+                status=200,
+                headers={
+                    "Content-Type": "application/rss+xml",
+                    "Cache-Control": "public, max-age=600",
+                    "ETag": '"feed-1"',
+                },
+                body=RSS,
+            ),
+        ):
+            document = self.provider.fetch("https://example.test/rss")
+
+        self.assertEqual('"feed-1"', document.etag)
+        self.assertIsNotNone(document.acquired_at)
+        self.assertEqual(600, int((document.fresh_until - document.acquired_at).total_seconds()))
+        self.assertEqual(len(RSS), document.metadata["byte_count"])
+
+    def test_render_filters_entries_by_explicit_visibility(self) -> None:
+        request = FeedRenderRequest(
+            format="atom",
+            title="Public updates",
+            feed_url="https://example.test/feed.atom",
+            home_url="https://example.test/",
+            entries=(
+                FeedEntry(id="public-1", title="Public", visibility="public"),
+                FeedEntry(id="tenant-1", title="Tenant", visibility="tenant"),
+            ),
+            allowed_visibilities=frozenset({"public"}),
+        )
+        result = self.provider.render(request)
+        root = SafeET.fromstring(result.body)
+
+        self.assertEqual(1, result.included_entries)
+        self.assertEqual(1, result.excluded_entries)
+        self.assertIn(b"Public", result.body)
+        self.assertNotIn(b"Tenant", result.body)
+        self.assertTrue(root.tag.endswith("feed"))
+
+    def test_unsafe_xml_is_rejected(self) -> None:
+        payload = b'<!DOCTYPE x [<!ENTITY y SYSTEM "file:///etc/passwd">]><rss>&y;</rss>'
+        with self.assertRaisesRegex(FeedCapabilityError, "not safe or valid"):
+            self.provider.parse(payload, source_url="https://example.test/rss")
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_manifest.py

@@ -0,0 +1,39 @@
+from __future__ import annotations
+
+import unittest
+
+from govoplan_core.core.tabular_sources import (
+    CAPABILITY_CONNECTORS_TABULAR_SNAPSHOT_WRITER,
+    CAPABILITY_CONNECTORS_TABULAR_SOURCES,
+)
+from govoplan_core.core.sanctions import (
+    CAPABILITY_CONNECTORS_SANCTIONS_SNAPSHOTS,
+)
+from govoplan_connectors.backend.manifest import manifest
+
+
+class ConnectorsManifestTests(unittest.TestCase):
+    def test_manifest_exposes_versioned_tabular_capabilities(self) -> None:
+        self.assertEqual("connectors", manifest.id)
+        self.assertIn("access", manifest.optional_dependencies)
+        self.assertIn(
+            "connectors.tabular_sources",
+            {interface.name for interface in manifest.provides_interfaces},
+        )
+        self.assertIn(
+            CAPABILITY_CONNECTORS_TABULAR_SOURCES,
+            manifest.capability_factories,
+        )
+        self.assertIn(
+            CAPABILITY_CONNECTORS_TABULAR_SNAPSHOT_WRITER,
+            manifest.capability_factories,
+        )
+        self.assertIn(
+            CAPABILITY_CONNECTORS_SANCTIONS_SNAPSHOTS,
+            manifest.capability_factories,
+        )
+        self.assertIsNotNone(manifest.migration_spec)
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_migrations.py

@@ -0,0 +1,47 @@
+from __future__ import annotations
+
+import tempfile
+import unittest
+from pathlib import Path
+
+from alembic.runtime.migration import MigrationContext
+from sqlalchemy import create_engine, inspect
+
+from govoplan_connectors.backend.manifest import get_manifest
+from govoplan_core.db.migrations import migrate_database
+
+
+class ConnectorsMigrationTests(unittest.TestCase):
+    def test_baseline_creates_connector_tables_and_head(self) -> None:
+        with tempfile.TemporaryDirectory(prefix="govoplan-connectors-migration-") as directory:
+            url = f"sqlite:///{Path(directory) / 'connectors.db'}"
+            migrate_database(
+                database_url=url,
+                enabled_modules=("connectors",),
+                manifest_factories=(get_manifest,),
+            )
+            engine = create_engine(url)
+            try:
+                with engine.connect() as connection:
+                    self.assertIn(
+                        "f7c8d9e0a1b2",
+                        set(MigrationContext.configure(connection).get_current_heads()),
+                    )
+                    self.assertIn(
+                        "connector_tabular_sources",
+                        inspect(connection).get_table_names(),
+                    )
+                    self.assertIn(
+                        "connector_sanctions_snapshots",
+                        inspect(connection).get_table_names(),
+                    )
+                    self.assertIn(
+                        "connector_sanctions_acquisition_runs",
+                        inspect(connection).get_table_names(),
+                    )
+            finally:
+                engine.dispose()
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_provider_state.py

@@ -0,0 +1,118 @@
+from __future__ import annotations
+
+from datetime import UTC, datetime
+import unittest
+
+from sqlalchemy import create_engine
+from sqlalchemy.orm import sessionmaker
+
+from govoplan_connectors.backend.db.models import (
+    ConnectorSanctionsAcquisitionRun,
+    ConnectorSanctionsSnapshot,
+    ConnectorTabularSource,
+)
+from govoplan_connectors.backend.manifest import manifest
+from govoplan_connectors.backend.provider_state import (
+    SANCTIONS_PROVIDER_ID,
+    TABULAR_PROVIDER_ID,
+    sanctions_provider_states,
+    tabular_provider_states,
+)
+from govoplan_core.core.provider_governance import ExternalProviderStateContext
+from govoplan_core.db.base import Base
+
+
+class ConnectorsProviderStateTests(unittest.TestCase):
+    def setUp(self) -> None:
+        self.engine = create_engine("sqlite+pysqlite:///:memory:", future=True)
+        Base.metadata.create_all(
+            self.engine,
+            tables=(
+                ConnectorTabularSource.__table__,
+                ConnectorSanctionsAcquisitionRun.__table__,
+                ConnectorSanctionsSnapshot.__table__,
+            ),
+        )
+        self.session = sessionmaker(bind=self.engine, expire_on_commit=False)()
+
+    def tearDown(self) -> None:
+        self.session.close()
+        self.engine.dispose()
+
+    def test_immutable_tabular_snapshot_reports_ready_state(self) -> None:
+        self.session.add(
+            ConnectorTabularSource(
+                id="tabular-1",
+                tenant_id="tenant-1",
+                source_name="monthly",
+                name="Monthly input",
+                status="active",
+                schema_version=1,
+                schema_=[{"name": "id", "type": "string"}],
+                rows=[{"id": "1"}],
+                fingerprint="a" * 64,
+                row_count=1,
+                byte_count=10,
+            )
+        )
+        self.session.commit()
+
+        state = tabular_provider_states(
+            ExternalProviderStateContext(session=self.session, tenant_id="tenant-1")
+        )[0]
+
+        self.assertEqual("healthy", state.health)
+        self.assertEqual("ready", state.recovery)
+        self.assertEqual("not_applicable", state.freshness)
+
+    def test_sanctions_state_hashes_binding_and_manifest_registers_state(self) -> None:
+        now = datetime.now(UTC)
+        run = ConnectorSanctionsAcquisitionRun(
+            id="run-1",
+            tenant_id="tenant-1",
+            provider_id="eu",
+            source_id="secret-source-name",
+            status="succeeded",
+            attempt_count=1,
+            started_at=now,
+            finished_at=now,
+        )
+        snapshot = ConnectorSanctionsSnapshot(
+            id="snapshot-1",
+            tenant_id="tenant-1",
+            provider_id="eu",
+            publisher="European Union",
+            jurisdiction="EU",
+            list_type="sanctions",
+            source_id="secret-source-name",
+            source_version="2026-08-01",
+            acquired_at=now,
+            source_url="https://source.example.test/list.xml",
+            content_type="application/xml",
+            byte_count=8,
+            sha256="b" * 64,
+            parser_version="1",
+            connector_run_id=run.id,
+            raw_content=b"<list/>",
+        )
+        run.snapshot_id = snapshot.id
+        self.session.add_all((run, snapshot))
+        self.session.commit()
+
+        state = sanctions_provider_states(
+            ExternalProviderStateContext(session=self.session, tenant_id="tenant-1")
+        )[0]
+
+        self.assertEqual("healthy", state.health)
+        self.assertEqual("ready", state.recovery)
+        rendered = str(state.to_dict())
+        self.assertNotIn("secret-source-name", rendered)
+        self.assertNotIn("source.example.test", rendered)
+        self.assertEqual(
+            {TABULAR_PROVIDER_ID, SANCTIONS_PROVIDER_ID},
+            {item.provider_id for item in manifest.external_provider_state_providers},
+        )
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_recovery.py

@@ -0,0 +1,263 @@
+from __future__ import annotations
+
+from datetime import datetime, timedelta, timezone
+import unittest
+from unittest.mock import patch
+
+from sqlalchemy import create_engine, select
+from sqlalchemy.orm import Session
+
+from govoplan_connectors.backend.db.models import ConnectorTabularSource
+from govoplan_connectors.backend.feeds import ConnectorFeedProvider
+from govoplan_connectors.backend.recovery import (
+    CONNECTOR_RECOVERY_OPERATIONS,
+    ConnectorRecoveryError,
+    begin_connector_external_mutation,
+    begin_connector_read_snapshot,
+)
+from govoplan_connectors.backend.router import api_import_feed_snapshot
+from govoplan_connectors.backend.schemas import FeedImportRequest
+from govoplan_connectors.backend.tabular_sources import WRITE_SCOPE
+from govoplan_core.auth import ApiPrincipal
+from govoplan_core.core.access import PrincipalRef
+from govoplan_core.core.recovery import (
+    RecoveryCheckpoint,
+    RecoveryOperation,
+    RecoveryStatus,
+)
+from govoplan_core.core.recovery_runtime import (
+    RecoveryOperationStateConflict,
+    claim_durable_recovery_operation,
+)
+from govoplan_core.core.tabular_sources import TabularSnapshotInput
+from govoplan_core.core.runtime_coordination import (
+    DistributedLease,
+    RuntimeIdentity,
+    bind_process_runtime_identity,
+)
+from govoplan_core.db.base import Base
+from govoplan_connectors.backend.tabular_sources import SqlTabularSourceProvider
+
+
+RSS = b"""<?xml version="1.0"?>
+<rss version="2.0"><channel><title>Updates</title>
+<link>https://example.test/</link><description>Updates</description>
+<item><guid>1</guid><title>One</title></item></channel></rss>"""
+
+
+def _identity(node: str, incarnation: str) -> RuntimeIdentity:
+    return RuntimeIdentity(
+        installation_id="connector-recovery-tests",
+        node_id=node,
+        incarnation=incarnation,
+        role="worker",
+        software_version="test",
+        composition_hash="a" * 64,
+    )
+
+
+def _principal() -> ApiPrincipal:
+    return ApiPrincipal(
+        principal=PrincipalRef(
+            account_id="account-1",
+            membership_id="membership-1",
+            tenant_id="tenant-1",
+            scopes=frozenset({WRITE_SCOPE}),
+        ),
+        account=object(),
+        user=object(),
+    )
+
+
+class ConnectorRecoveryTests(unittest.TestCase):
+    def setUp(self) -> None:
+        self.engine = create_engine("sqlite+pysqlite:///:memory:")
+        Base.metadata.create_all(
+            self.engine,
+            tables=(
+                DistributedLease.__table__,
+                RecoveryOperation.__table__,
+                RecoveryCheckpoint.__table__,
+                ConnectorTabularSource.__table__,
+            ),
+        )
+        self.session = Session(self.engine, expire_on_commit=False)
+        bind_process_runtime_identity(_identity("node-1", "incarnation-1"))
+
+    def tearDown(self) -> None:
+        bind_process_runtime_identity(None)
+        self.session.close()
+        self.engine.dispose()
+
+    def test_feed_snapshot_and_recovery_checkpoint_commit_atomically_and_replay(self) -> None:
+        document = ConnectorFeedProvider().parse(
+            RSS,
+            source_url="https://example.test/feed.xml",
+        )
+        payload = FeedImportRequest(
+            url="https://example.test/feed.xml",
+            name="Updates",
+            source_name="updates",
+        )
+        with (
+            patch(
+                "govoplan_connectors.backend.router.feed_transport.fetch",
+                return_value=document,
+            ) as fetch,
+            patch("govoplan_connectors.backend.router.audit_event"),
+        ):
+            first = api_import_feed_snapshot(
+                payload,
+                session=self.session,
+                principal=_principal(),
+                idempotency_key="feed-import-1",
+            )
+            replay = api_import_feed_snapshot(
+                payload,
+                session=self.session,
+                principal=_principal(),
+                idempotency_key="feed-import-1",
+            )
+
+        self.assertEqual(first.ref, replay.ref)
+        fetch.assert_called_once()
+        operation = self.session.scalar(select(RecoveryOperation))
+        assert operation is not None
+        self.assertEqual(RecoveryStatus.SUCCEEDED.value, operation.status)
+        self.assertEqual(
+            first.ref.removeprefix("snapshot:"),
+            operation.resource_id,
+        )
+
+    def test_recovery_metadata_distinguishes_reads_from_external_mutations(self) -> None:
+        declarations = {
+            item.operation_type: item for item in CONNECTOR_RECOVERY_OPERATIONS
+        }
+
+        self.assertFalse(declarations["read-snapshot"].provider_mutation)
+        self.assertTrue(declarations["read-snapshot"].implemented)
+        self.assertTrue(declarations["external-mutation"].provider_mutation)
+        self.assertFalse(declarations["external-mutation"].implemented)
+
+    def test_stale_atomic_connector_fence_fails_without_claiming_an_effect(self) -> None:
+        recovery = begin_connector_read_snapshot(
+            self.session,
+            tenant_id="tenant-1",
+            provider_id="provider-1",
+            idempotency_key="read-1",
+            source_revision="revision-1",
+            cursor="cursor-1",
+            dry_run_evidence={"performed": True, "approved": True},
+        )
+        lease = self.session.scalar(select(DistributedLease))
+        assert lease is not None
+        lease.expires_at = datetime.now(timezone.utc) - timedelta(seconds=1)
+        self.session.commit()
+        bind_process_runtime_identity(_identity("node-2", "incarnation-2"))
+
+        with self.assertRaises(RecoveryOperationStateConflict):
+            claim_durable_recovery_operation(
+                recovery.operation.session_factory,
+                identity=_identity("node-2", "incarnation-2"),
+                operation_id=recovery.operation_id,
+            )
+
+        operation = self.session.get(RecoveryOperation, recovery.operation_id)
+        self.session.refresh(operation)
+        self.assertEqual(RecoveryStatus.FAILED.value, operation.status)
+
+    def test_external_mutation_unknown_outcome_blocks_blind_retry(self) -> None:
+        kwargs = {
+            "tenant_id": "tenant-1",
+            "provider_id": "provider-1",
+            "idempotency_key": "publish-1",
+            "request_sha256": "b" * 64,
+            "source_revision": "revision-1",
+            "cursor": None,
+            "dry_run_evidence": {"performed": True, "approved": True},
+            "resource_type": "external_record",
+            "resource_id": "record-1",
+        }
+        recovery = begin_connector_external_mutation(self.session, **kwargs)
+        recovery.outcome_unknown(
+            summary="The provider connection closed after dispatch",
+            provider_code="connection_closed",
+        )
+
+        with self.assertRaises(ConnectorRecoveryError):
+            begin_connector_external_mutation(self.session, **kwargs)
+        operation = self.session.get(RecoveryOperation, recovery.operation_id)
+        self.session.refresh(operation)
+        self.assertEqual(RecoveryStatus.OUTCOME_UNKNOWN.value, operation.status)
+
+    def test_tampered_chain_rolls_back_the_atomic_connector_projection(self) -> None:
+        recovery = begin_connector_read_snapshot(
+            self.session,
+            tenant_id="tenant-1",
+            provider_id="provider-1",
+            idempotency_key="tampered-read",
+            source_revision=None,
+            cursor=None,
+            dry_run_evidence={"performed": False, "reason": "read-only"},
+        )
+        checkpoint = self.session.scalar(
+            select(RecoveryCheckpoint)
+            .where(RecoveryCheckpoint.operation_id == recovery.operation_id)
+            .order_by(RecoveryCheckpoint.sequence)
+            .limit(1)
+        )
+        assert checkpoint is not None
+        checkpoint.summary = "tampered"
+        self.session.commit()
+        source = SqlTabularSourceProvider().create_snapshot(
+            self.session,
+            _principal(),
+            snapshot=TabularSnapshotInput(
+                name="Tampered",
+                source_name="tampered",
+                rows=({"id": 1},),
+            ),
+            source_id=recovery.resource_id,
+        )
+
+        with self.assertRaises(ConnectorRecoveryError):
+            recovery.commit_success(
+                self.session,
+                evidence={
+                    "verified": True,
+                    "checks": {"snapshot_ref": source.ref},
+                },
+            )
+
+        self.assertIsNone(
+            self.session.get(ConnectorTabularSource, recovery.resource_id)
+        )
+        operation = self.session.get(RecoveryOperation, recovery.operation_id)
+        self.session.refresh(operation)
+        self.assertEqual(RecoveryStatus.RUNNING.value, operation.status)
+
+    def test_definitive_external_rejection_is_terminal(self) -> None:
+        recovery = begin_connector_external_mutation(
+            self.session,
+            tenant_id="tenant-1",
+            provider_id="provider-1",
+            idempotency_key="publish-rejected",
+            request_sha256="c" * 64,
+            source_revision="revision-1",
+            cursor=None,
+            dry_run_evidence={"performed": True, "approved": True},
+            resource_type="external_record",
+            resource_id="record-2",
+        )
+        recovery.reject(
+            summary="The provider rejected the requested revision",
+            provider_code="revision_conflict",
+        )
+
+        operation = self.session.get(RecoveryOperation, recovery.operation_id)
+        self.session.refresh(operation)
+        self.assertEqual(RecoveryStatus.REJECTED.value, operation.status)
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_sanctions_sources.py

@@ -0,0 +1,391 @@
+from __future__ import annotations
+
+from datetime import timedelta
+import hashlib
+import unittest
+from unittest.mock import patch
+from urllib.error import URLError
+
+from sqlalchemy import create_engine
+from sqlalchemy.orm import Session
+
+from govoplan_connectors.backend.db.models import (
+    ConnectorSanctionsAcquisitionRun,
+    ConnectorSanctionsSnapshot,
+)
+from govoplan_connectors.backend.sanctions_sources import (
+    SANCTIONS_READ_SCOPE,
+    SANCTIONS_REFRESH_SCOPE,
+    SOURCE_DEFINITIONS,
+    SYNTHETIC_PROVIDER_ID,
+    SYNTHETIC_UN_XML,
+    SanctionsSourceError,
+    SqlSanctionsSnapshotProvider,
+    TransportResponse,
+    UNSC_PROVIDER_ID,
+    UrllibSanctionsTransport,
+)
+from govoplan_core.auth import ApiPrincipal
+from govoplan_core.core.access import PrincipalRef
+from govoplan_core.core.recovery import (
+    RecoveryCheckpoint,
+    RecoveryOperation,
+    RecoveryStatus,
+)
+from govoplan_core.core.runtime_coordination import (
+    DistributedLease,
+    RuntimeIdentity,
+    bind_process_runtime_identity,
+)
+from govoplan_core.db.base import Base, utcnow
+
+
+def principal(
+    tenant_id: str = "tenant-1",
+    *,
+    scopes: tuple[str, ...] = (
+        SANCTIONS_READ_SCOPE,
+        SANCTIONS_REFRESH_SCOPE,
+    ),
+) -> ApiPrincipal:
+    return ApiPrincipal(
+        principal=PrincipalRef(
+            account_id="account-1",
+            membership_id="membership-1",
+            tenant_id=tenant_id,
+            scopes=frozenset(scopes),
+        ),
+        account=object(),
+        user=object(),
+    )
+
+
+class _Transport:
+    def __init__(self, responses):
+        self.responses = list(responses)
+        self.headers = []
+
+    def fetch(self, definition, *, headers):
+        del definition
+        self.headers.append(dict(headers))
+        response = self.responses.pop(0)
+        if isinstance(response, Exception):
+            raise response
+        return response
+
+
+def response(
+    content: bytes = SYNTHETIC_UN_XML,
+    *,
+    status: int = 200,
+    content_type: str = "application/xml",
+    etag: str = '"fixture-v1"',
+) -> TransportResponse:
+    return TransportResponse(
+        status=status,
+        final_url="https://scsanctions.un.org/consolidated.xml",
+        headers={
+            "content-type": content_type,
+            "etag": etag,
+        },
+        content=content,
+        attempts=1,
+    )
+
+
+class SanctionsSourcesTests(unittest.TestCase):
+    def setUp(self) -> None:
+        self.engine = create_engine("sqlite:///:memory:")
+        Base.metadata.create_all(
+            self.engine,
+            tables=(
+                DistributedLease.__table__,
+                RecoveryOperation.__table__,
+                RecoveryCheckpoint.__table__,
+                ConnectorSanctionsAcquisitionRun.__table__,
+                ConnectorSanctionsSnapshot.__table__,
+            ),
+        )
+        self.session = Session(self.engine)
+        bind_process_runtime_identity(
+            RuntimeIdentity(
+                installation_id="connectors-tests",
+                node_id="connectors-test-node",
+                incarnation="connectors-test-incarnation",
+                role="worker",
+                software_version="test",
+                composition_hash="a" * 64,
+            )
+        )
+
+    def tearDown(self) -> None:
+        bind_process_runtime_identity(None)
+        self.session.close()
+        self.engine.dispose()
+
+    def test_fixture_refreshes_are_immutable_and_evidence_is_readable(
+        self,
+    ) -> None:
+        provider = SqlSanctionsSnapshotProvider()
+
+        first = provider.refresh_source(
+            self.session,
+            principal(),
+            provider_id=SYNTHETIC_PROVIDER_ID,
+        )
+        second = provider.refresh_source(
+            self.session,
+            principal(),
+            provider_id=SYNTHETIC_PROVIDER_ID,
+        )
+        self.session.commit()
+
+        self.assertEqual("succeeded", first.status)
+        self.assertEqual("succeeded", second.status)
+        self.assertNotEqual(first.snapshot.ref, second.snapshot.ref)
+        self.assertEqual(
+            hashlib.sha256(SYNTHETIC_UN_XML).hexdigest(),
+            first.snapshot.sha256,
+        )
+        self.assertEqual(
+            SYNTHETIC_UN_XML,
+            provider.read_snapshot(
+                self.session,
+                principal(),
+                snapshot_ref=first.snapshot.ref,
+            ).content,
+        )
+        runs = provider.list_runs(
+            self.session,
+            principal(),
+        )
+        self.assertTrue(
+            all(
+                run.request_evidence["subject_data_transmitted"]
+                is False
+                for run in runs
+            )
+        )
+
+    def test_conditional_fetch_reuses_prior_immutable_snapshot(self) -> None:
+        transport = _Transport(
+            (
+                response(),
+                response(b"", status=304),
+            )
+        )
+        provider = SqlSanctionsSnapshotProvider(transport)
+        first = provider.refresh_source(
+            self.session,
+            principal(),
+            provider_id=UNSC_PROVIDER_ID,
+        )
+        second = provider.refresh_source(
+            self.session,
+            principal(),
+            provider_id=UNSC_PROVIDER_ID,
+        )
+
+        self.assertEqual("not_modified", second.status)
+        self.assertEqual(first.snapshot.ref, second.snapshot.ref)
+        self.assertEqual(
+            {'If-None-Match': '"fixture-v1"'},
+            transport.headers[1],
+        )
+        self.assertEqual(
+            1,
+            self.session.query(ConnectorSanctionsSnapshot).count(),
+        )
+
+    def test_malformed_and_changed_sources_have_explicit_health(
+        self,
+    ) -> None:
+        cases = (
+            (
+                b"<CONSOLIDATED_LIST>",
+                "application/xml",
+                "malformed",
+            ),
+            (
+                b"<DIFFERENT><INDIVIDUALS/><ENTITIES/></DIFFERENT>",
+                "application/xml",
+                "unexpected_change",
+            ),
+            (
+                SYNTHETIC_UN_XML,
+                "text/html",
+                "unexpected_change",
+            ),
+        )
+        for payload, content_type, expected in cases:
+            with self.subTest(expected=expected, content_type=content_type):
+                provider = SqlSanctionsSnapshotProvider(
+                    _Transport(
+                        (
+                            response(
+                                payload,
+                                content_type=content_type,
+                            ),
+                        )
+                    )
+                )
+                result = provider.refresh_source(
+                    self.session,
+                    principal(),
+                    provider_id=UNSC_PROVIDER_ID,
+                )
+                self.assertEqual(expected, result.status)
+                self.assertIsNotNone(result.error)
+
+    def test_unavailable_source_becomes_stale_when_evidence_is_old(
+        self,
+    ) -> None:
+        provider = SqlSanctionsSnapshotProvider(
+            _Transport(
+                (
+                    response(),
+                    SanctionsSourceError("offline"),
+                )
+            )
+        )
+        first = provider.refresh_source(
+            self.session,
+            principal(),
+            provider_id=UNSC_PROVIDER_ID,
+        )
+        record = self.session.get(
+            ConnectorSanctionsSnapshot,
+            first.snapshot.ref.removeprefix("sanctions-snapshot:"),
+        )
+        record.acquired_at = utcnow() - timedelta(days=3)
+        self.session.commit()
+
+        result = provider.refresh_source(
+            self.session,
+            principal(),
+            provider_id=UNSC_PROVIDER_ID,
+        )
+
+        self.assertEqual("stale", result.status)
+        self.assertEqual(first.snapshot.ref, result.snapshot.ref)
+
+    def test_idempotent_refresh_replays_the_committed_acquisition(self) -> None:
+        transport = _Transport((response(),))
+        provider = SqlSanctionsSnapshotProvider(transport)
+
+        first = provider.refresh_source(
+            self.session,
+            principal(),
+            provider_id=UNSC_PROVIDER_ID,
+            idempotency_key="scheduled-refresh-1",
+        )
+        replay = provider.refresh_source(
+            self.session,
+            principal(),
+            provider_id=UNSC_PROVIDER_ID,
+            idempotency_key="scheduled-refresh-1",
+        )
+
+        self.assertEqual(first.run_id, replay.run_id)
+        self.assertEqual(first.snapshot.ref, replay.snapshot.ref)
+        self.assertEqual([], transport.responses)
+        operation = self.session.query(RecoveryOperation).one()
+        self.assertEqual(RecoveryStatus.SUCCEEDED.value, operation.status)
+
+    def test_provider_failure_commits_failed_run_and_terminal_recovery(self) -> None:
+        provider = SqlSanctionsSnapshotProvider(
+            _Transport((SanctionsSourceError("offline"),))
+        )
+
+        result = provider.refresh_source(
+            self.session,
+            principal(),
+            provider_id=UNSC_PROVIDER_ID,
+        )
+
+        self.assertEqual("unavailable", result.status)
+        operation = self.session.query(RecoveryOperation).one()
+        self.assertEqual(RecoveryStatus.FAILED.value, operation.status)
+        self.assertEqual(
+            result.run_id,
+            self.session.query(ConnectorSanctionsAcquisitionRun).one().id,
+        )
+
+    def test_snapshot_access_is_tenant_and_scope_isolated(self) -> None:
+        provider = SqlSanctionsSnapshotProvider()
+        created = provider.refresh_source(
+            self.session,
+            principal(),
+            provider_id=SYNTHETIC_PROVIDER_ID,
+        )
+
+        self.assertIsNone(
+            provider.get_snapshot(
+                self.session,
+                principal("tenant-2"),
+                snapshot_ref=created.snapshot.ref,
+            )
+        )
+        with self.assertRaisesRegex(Exception, "Missing scope"):
+            provider.list_snapshots(
+                self.session,
+                principal(scopes=()),
+            )
+
+    def test_transport_retries_transient_network_failures(self) -> None:
+        class _Headers(dict):
+            pass
+
+        class _Response:
+            status = 200
+            headers = _Headers(
+                {
+                    "Content-Type": "application/xml",
+                    "Content-Length": str(len(SYNTHETIC_UN_XML)),
+                }
+            )
+
+            def __enter__(self):
+                return self
+
+            def __exit__(self, *args):
+                return False
+
+            def geturl(self):
+                return (
+                    "https://scsanctions.un.org/"
+                    "resources/xml/en/consolidated.xml"
+                )
+
+            def read(self, size):
+                del size
+                if hasattr(self, "_read"):
+                    return b""
+                self._read = True
+                return SYNTHETIC_UN_XML
+
+        opener = unittest.mock.Mock()
+        opener.open.side_effect = (
+            URLError("temporary"),
+            URLError("temporary"),
+            _Response(),
+        )
+        sleeps = []
+        transport = UrllibSanctionsTransport(
+            sleeper=sleeps.append
+        )
+        with patch(
+            "govoplan_connectors.backend.sanctions_sources.build_opener",
+            return_value=opener,
+        ):
+            fetched = transport.fetch(
+                SOURCE_DEFINITIONS[UNSC_PROVIDER_ID],
+                headers={},
+            )
+
+        self.assertEqual(3, fetched.attempts)
+        self.assertEqual([1.0, 2.0], sleeps)
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_tabular_sources.py

@@ -0,0 +1,268 @@
+from __future__ import annotations
+
+import unittest
+
+from fastapi import HTTPException
+from sqlalchemy import create_engine
+from sqlalchemy.orm import sessionmaker
+
+from govoplan_core.auth import ApiPrincipal
+from govoplan_core.core.access import PrincipalRef
+from govoplan_core.core.tabular_sources import (
+    TabularReadRequest,
+    TabularSnapshotInput,
+    TabularSourceAccessError,
+    TabularSourceUnavailableError,
+    TabularSourceValidationError,
+)
+from govoplan_core.db.base import Base
+from govoplan_connectors.backend.db.models import ConnectorTabularSource
+from govoplan_connectors.backend.router import api_create_tabular_snapshot
+from govoplan_connectors.backend.schemas import SnapshotCreateRequest
+from govoplan_connectors.backend.tabular_sources import (
+    READ_SCOPE,
+    WRITE_SCOPE,
+    SqlTabularSourceProvider,
+    parse_csv_snapshot,
+)
+
+
+def principal(
+    tenant_id: str = "tenant-1",
+    *,
+    scopes: tuple[str, ...] = (READ_SCOPE, WRITE_SCOPE),
+) -> ApiPrincipal:
+    return ApiPrincipal(
+        principal=PrincipalRef(
+            account_id="account-1",
+            membership_id="membership-1",
+            tenant_id=tenant_id,
+            scopes=frozenset(scopes),
+        ),
+        account=object(),
+        user=object(),
+    )
+
+
+class ConnectorsTabularSourceTests(unittest.TestCase):
+    def setUp(self) -> None:
+        self.engine = create_engine("sqlite:///:memory:")
+        Base.metadata.create_all(self.engine, tables=[ConnectorTabularSource.__table__])
+        self.Session = sessionmaker(bind=self.engine)
+        self.session = self.Session()
+        self.provider = SqlTabularSourceProvider()
+
+    def tearDown(self) -> None:
+        self.session.close()
+        Base.metadata.drop_all(self.engine, tables=[ConnectorTabularSource.__table__])
+        self.engine.dispose()
+
+    def test_snapshot_round_trip_preserves_schema_fingerprint_and_bounds(self) -> None:
+        created = self.provider.create_snapshot(
+            self.session,
+            principal(),
+            snapshot=TabularSnapshotInput(
+                name="Monthly cases",
+                source_name="monthly_cases_2026_07",
+                rows=(
+                    {"case_id": "A-1", "amount": 12, "active": True},
+                    {"case_id": "A-2", "amount": None, "active": False},
+                ),
+            ),
+        )
+        self.session.commit()
+
+        listed = self.provider.list_sources(self.session, principal())
+        preview = self.provider.read_source(
+            self.session,
+            principal(),
+            request=TabularReadRequest(
+                source_ref=created.ref,
+                limit=1,
+                expected_fingerprint=created.fingerprint,
+            ),
+        )
+
+        self.assertEqual((created.ref,), tuple(source.ref for source in listed))
+        self.assertEqual(
+            ["case_id", "amount", "active"],
+            [column.name for column in created.schema],
+        )
+        self.assertEqual(2, preview.total_rows)
+        self.assertEqual(1, len(preview.rows))
+        self.assertTrue(preview.truncated)
+        self.assertEqual(created.fingerprint, preview.source.fingerprint)
+        self.assertEqual("cached", preview.source.source_mode)
+        self.assertTrue(preview.source.pushdown.projections)
+        self.assertTrue(preview.source.pushdown.pagination)
+        self.assertEqual("healthy", preview.source.health.status)
+        self.assertGreater(preview.returned_bytes, 2)
+        self.assertEqual(1, preview.effective_row_limit)
+        self.assertEqual("preview.row_limit_reached", preview.diagnostics[0].code)
+
+    def test_preview_enforces_byte_time_and_provider_ceiling_budgets(self) -> None:
+        created = self.provider.create_snapshot(
+            self.session,
+            principal(),
+            snapshot=TabularSnapshotInput(
+                name="Bounded",
+                source_name="bounded",
+                rows=(
+                    {"id": 1, "value": "first"},
+                    {"id": 2, "value": "second"},
+                ),
+            ),
+        )
+        self.session.commit()
+
+        bounded = self.provider.read_source(
+            self.session,
+            principal(),
+            request=TabularReadRequest(
+                source_ref=created.ref,
+                limit=500,
+                max_bytes=35,
+                timeout_ms=2_000,
+            ),
+        )
+        self.assertEqual(1, len(bounded.rows))
+        self.assertTrue(bounded.truncated)
+        self.assertEqual(
+            "preview.byte_limit_reached",
+            bounded.diagnostics[-1].code,
+        )
+        with self.assertRaisesRegex(
+            TabularSourceValidationError,
+            "single source row exceeds",
+        ):
+            self.provider.read_source(
+                self.session,
+                principal(),
+                request=TabularReadRequest(
+                    source_ref=created.ref,
+                    max_bytes=2,
+                ),
+            )
+
+        tightened = self.provider.read_source(
+            self.session,
+            principal(),
+            request=TabularReadRequest(
+                source_ref=created.ref,
+                limit=5_000,
+                max_bytes=5_000_000,
+                timeout_ms=10_000,
+            ),
+        )
+        self.assertEqual(500, tightened.effective_row_limit)
+        self.assertEqual(1_000_000, tightened.effective_byte_limit)
+        self.assertEqual(2_000, tightened.effective_timeout_ms)
+        self.assertEqual(
+            {
+                "preview.row_limit_tightened",
+                "preview.byte_limit_tightened",
+                "preview.timeout_tightened",
+            },
+            {item.code for item in tightened.diagnostics},
+        )
+
+        times = iter((0.0, 0.01))
+        timeout_provider = SqlTabularSourceProvider(clock=lambda: next(times))
+        with self.assertRaisesRegex(
+            TabularSourceUnavailableError,
+            "time budget",
+        ):
+            timeout_provider.read_source(
+                self.session,
+                principal(),
+                request=TabularReadRequest(
+                    source_ref=created.ref,
+                    timeout_ms=1,
+                ),
+            )
+
+    def test_tenant_and_scope_isolation_are_enforced(self) -> None:
+        created = self.provider.create_snapshot(
+            self.session,
+            principal(),
+            snapshot=TabularSnapshotInput(
+                name="Private",
+                source_name="private_source",
+                rows=({"id": 1},),
+            ),
+        )
+        self.session.commit()
+
+        self.assertEqual((), self.provider.list_sources(self.session, principal("tenant-2")))
+        self.assertIsNone(
+            self.provider.get_source(
+                self.session,
+                principal("tenant-2"),
+                source_ref=created.ref,
+            )
+        )
+        with self.assertRaises(TabularSourceAccessError):
+            self.provider.list_sources(
+                self.session,
+                principal(scopes=()),
+            )
+
+    def test_duplicate_source_name_and_stale_fingerprint_are_rejected(self) -> None:
+        snapshot = TabularSnapshotInput(
+            name="Cases",
+            source_name="cases",
+            rows=({"id": 1},),
+        )
+        created = self.provider.create_snapshot(self.session, principal(), snapshot=snapshot)
+        self.session.commit()
+
+        with self.assertRaises(TabularSourceValidationError):
+            self.provider.create_snapshot(self.session, principal(), snapshot=snapshot)
+        with self.assertRaises(TabularSourceValidationError):
+            self.provider.read_source(
+                self.session,
+                principal(),
+                request=TabularReadRequest(
+                    source_ref=created.ref,
+                    expected_fingerprint="stale",
+                ),
+            )
+
+    def test_csv_parser_infers_scalar_values_and_rejects_duplicate_headers(self) -> None:
+        rows = parse_csv_snapshot(
+            "\ufeffid;amount;active;note\n0012;12.5;true;\n2;7;false;ok\n",
+            delimiter=";",
+        )
+
+        self.assertEqual(
+            (
+                {"id": "0012", "amount": 12.5, "active": True, "note": None},
+                {"id": 2, "amount": 7, "active": False, "note": "ok"},
+            ),
+            rows,
+        )
+        with self.assertRaises(TabularSourceValidationError):
+            parse_csv_snapshot("id,id\n1,2\n", delimiter=",")
+        with self.assertRaises(TabularSourceValidationError):
+            parse_csv_snapshot("id,name\n1,Ada,extra\n", delimiter=",")
+
+    def test_malformed_csv_api_request_is_reported_as_validation_error(self) -> None:
+        payload = SnapshotCreateRequest(
+            name="Malformed",
+            source_name="malformed",
+            format="csv",
+            csv_text="id,name\n1,Ada,extra\n",
+        )
+
+        with self.assertRaises(HTTPException) as raised:
+            api_create_tabular_snapshot(
+                payload,
+                session=self.session,
+                principal=principal(),
+            )
+
+        self.assertEqual(422, raised.exception.status_code)
+
+
+if __name__ == "__main__":
+    unittest.main()