feat: add governed Files DSAR coverage

README.md

@@ -55,6 +55,15 @@ for example for managed attachment selection and campaign file sharing. Files
 does not import campaign internals; campaign share/existence checks use the core
 `campaigns.access` capability registered by the campaign module.
 
+Files also publishes the optional `privacy.dsar.files` capability. The Core
+data-subject workflow can use it to collect bounded, tenant-scoped file,
+version, share, folder, evidence, and non-secret configuration metadata for a
+direct membership subject. The provider may revoke a subject-targeted share or
+detach mutable actor references idempotently, but it never exports credential
+material or raw bytes and never bypasses Files retention, legal hold, evidence,
+purge approval, audit, or recovery controls. See the handbook's data-subject
+request coverage section for the review and erasure boundary.
+
 Platform RBAC and governance rules are documented in `govoplan-core/docs/`.
 
 Managed files can carry source provenance for connector and import workflows.

docs/FILES_HANDBOOK.md

@@ -208,6 +208,39 @@ Hard purge is deliberately separate from ordinary delete:
 Automatic time-based purge scheduling is not implemented. Operators initiate
 preview, execute, and garbage collection under their local retention process.
 
+### Data-subject request coverage
+
+Files registers `privacy.dsar.files` when the module is active. The provider
+requires a corroborated tenant membership identifier (or a namespaced Files
+user reference), searches only that tenant, and fails explicitly if its bounded
+result limit would be exceeded. It reports managed assets, exact versions,
+folders, user-targeted shares, Form and Campaign evidence, connector
+configuration actor references, and integrity-operation evidence.
+
+The export includes only governed metadata. It never embeds file bytes, blob
+storage keys, passwords, tokens, environment-variable names, secret-provider
+references, or encrypted connector values. A reviewer follows the authorized
+version download route when the file itself must be inspected.
+
+The erasure plan deliberately separates four outcomes:
+
+- an active share aimed at the subject can be revoked idempotently;
+- mutable creator/updater references can be detached after tenant, subject, and
+  current-value revalidation;
+- legal hold, active retention, submitted Form evidence, Campaign delivery
+  evidence, connector configuration history, and integrity evidence are
+  retained with a reason; and
+- unstructured file content, ownership, filenames, and paths require manual
+  review.
+
+DSAR execution never invokes physical byte deletion. If the privacy decision
+authorizes erasure, the operator must use the separate Files soft-delete, purge
+preview, approval, execution, and blob-GC sequence. This preserves its distinct
+authority, evidence blockers, audit trail, distributed fencing, and recovery
+ledger semantics. Email, account, or identity selectors alone are insufficient
+because Files does not import the Access directory; the Access search supplies
+the corroborated membership reference for Files coverage.
+
 ### Find and download files
 
 Files can list by owner and path, use cursor pagination, and consume incremental
@@ -1018,6 +1051,7 @@ returning different content or credentials.
 | Organization | Folders, bulk rename preview/apply, move/copy, drag-and-drop, ZIP download, pattern resolution, and API restoration preserving versions/provenance | General file-history UI and user-driven append-version UI |
 | Sharing | User/group/tenant/campaign grants, expiry, idempotent revocation, searchable share-management UI, and campaign linkage display | Richer policy-driven share lifecycles |
 | Deletion/retention | Soft-delete and restore assets/folders/spaces; optimistic retention and legal-hold controls; preview-bound, approval-referenced hard purge; reference-checked blob GC; immediate audited connector-secret scrubbing | Automatic time-based purge scheduling and richer lifecycle administration UI |
+| Privacy requests | Tenant-scoped bounded DSAR metadata search; retained/manual/revoke/detach planning; idempotent share revocation and mutable actor-reference detachment; explicit separation from byte purge | Content-specific automated redaction and policy-specific approval remain manual or belong to the owning process |
 | Connector governance | Scoped profiles/credentials/policies, effective source explanation, separate credentials, linked user/group spaces | Provider-owned external secret lifecycle; API `secret_ref` remains rejected |
 | HTTP connectors | Pinned, bounded, no-redirect Seafile and WebDAV/Nextcloud browse/import/manual sync | Background/folder sync, remote mutation, long-running transfer workers |
 | SMB and S3 connectors | Provider descriptors, browse/import/manual sync, pinned SDK transports, redirect/retry/referral transport-contract tests, and explicit conditional S3 write-back with Core-ledger recovery | Live topology smoke evidence, provider-specific OAuth, additional provider writes, and background indexing remain separate deployment or connector-module concerns |
@@ -1053,7 +1087,10 @@ Before releasing Files:
     and applied orphan cleanup; inspect their `files` operations in Ops.
 11. Exercise a Form evidence grant, token replay, wrong-submission reference,
     expired token, unsupported media type, and quarantined-file rejection.
-12. Update the implemented/planned table whenever a boundary changes.
+12. Exercise a Files DSAR search/plan, repeat an approved reversible action, and
+    confirm retained file content can only be erased through the separate purge
+    authority and recovery path.
+13. Update the implemented/planned table whenever a boundary changes.
 
 ## Related documents
 

src/govoplan_files/backend/dsar_provider.py

@@ -0,0 +1,879 @@
+from __future__ import annotations
+
+from collections.abc import Sequence
+from datetime import datetime, timezone
+
+from sqlalchemy import inspect, or_
+from sqlalchemy.orm import Session
+
+from govoplan_core.core.dsar import (
+    DsarErasureActionRef,
+    DsarExecutionResultRef,
+    DsarRecordRef,
+    DsarSubjectRef,
+    dsar_capability_name,
+)
+from govoplan_files.backend.db.models import (
+    CampaignAttachmentUse,
+    FileAsset,
+    FileConnectorCredential,
+    FileConnectorPolicy,
+    FileConnectorProfile,
+    FileConnectorSpace,
+    FileFolder,
+    FileFormEvidenceGrant,
+    FileIntegrityFinding,
+    FileIntegrityScan,
+    FileShare,
+    FileVersion,
+)
+
+
+FILES_DSAR_CAPABILITY = dsar_capability_name("files")
+_MAX_RECORDS = 5_000
+
+
+class FilesDsarProvider:
+    provider_id = "files"
+    module_id = "files"
+
+    def search_subject(
+        self,
+        session: object,
+        *,
+        tenant_id: str,
+        subject: DsarSubjectRef,
+    ) -> Sequence[DsarRecordRef]:
+        db = _session(session)
+        subject_user_id = _subject_user_id(subject)
+        if subject_user_id is None:
+            return ()
+
+        records: list[DsarRecordRef] = []
+        form_evidence_available = _has_table(db, FileFormEvidenceGrant)
+        campaign_evidence_available = _has_table(db, CampaignAttachmentUse)
+        retention_reasons: dict[str, str | None] = {}
+
+        def retention_reason(asset: FileAsset) -> str | None:
+            if asset.id not in retention_reasons:
+                retention_reasons[asset.id] = _asset_retention_reason(
+                    db,
+                    asset,
+                    form_evidence_available=form_evidence_available,
+                    campaign_evidence_available=campaign_evidence_available,
+                )
+            return retention_reasons[asset.id]
+
+        def append(record: DsarRecordRef) -> None:
+            if len(records) >= _MAX_RECORDS:
+                raise ValueError(
+                    "Files DSAR match limit exceeded; narrow the subject selectors."
+                )
+            records.append(record)
+
+        assets = _bounded_rows(
+            db.query(FileAsset)
+            .filter(
+                FileAsset.tenant_id == tenant_id,
+                or_(
+                    FileAsset.owner_user_id == subject_user_id,
+                    FileAsset.created_by_user_id == subject_user_id,
+                ),
+            )
+            .order_by(FileAsset.id)
+        )
+        for asset in assets:
+            match_fields = _matching_fields(
+                asset,
+                subject_user_id,
+                ("owner_user_id", "created_by_user_id"),
+            )
+            asset_retention_reason = retention_reason(asset)
+            append(
+                _record(
+                    "file_asset",
+                    asset.id,
+                    "managed_file",
+                    asset.filename,
+                    {
+                        "match_fields": match_fields,
+                        "owner_type": asset.owner_type,
+                        "display_path": asset.display_path,
+                        "filename": asset.filename,
+                        "description": asset.description,
+                        "deleted_at": _iso(asset.deleted_at),
+                        "retained_until": _iso(asset.retained_until),
+                        "legal_hold": asset.legal_hold,
+                        "lifecycle_reason": asset.lifecycle_reason,
+                        "lifecycle_revision": asset.lifecycle_revision,
+                    },
+                    observed_at=asset.updated_at,
+                    immutable=asset_retention_reason is not None,
+                    retention_reason=asset_retention_reason,
+                    source_path=f"/files?file={asset.id}",
+                )
+            )
+
+        version_rows = _bounded_rows(
+            db.query(FileVersion, FileAsset)
+            .join(FileAsset, FileAsset.id == FileVersion.file_asset_id)
+            .filter(
+                FileVersion.tenant_id == tenant_id,
+                FileAsset.tenant_id == tenant_id,
+                or_(
+                    FileAsset.owner_user_id == subject_user_id,
+                    FileVersion.created_by_user_id == subject_user_id,
+                ),
+            )
+            .order_by(FileVersion.id)
+        )
+        for version, asset in version_rows:
+            asset_retention_reason = retention_reason(asset)
+            append(
+                _record(
+                    "file_version",
+                    version.id,
+                    "managed_file_version",
+                    version.filename_at_upload,
+                    {
+                        "match_fields": (
+                            ["asset.owner_user_id"]
+                            if asset.owner_user_id == subject_user_id
+                            else []
+                        )
+                        + (
+                            ["created_by_user_id"]
+                            if version.created_by_user_id == subject_user_id
+                            else []
+                        ),
+                        "file_asset_id": version.file_asset_id,
+                        "version_number": version.version_number,
+                        "filename_at_upload": version.filename_at_upload,
+                        "display_path_at_upload": version.display_path_at_upload,
+                        "content_type": version.content_type,
+                        "size_bytes": version.size_bytes,
+                        "checksum_sha256": version.checksum_sha256,
+                        "created_at": _iso(version.created_at),
+                    },
+                    observed_at=version.updated_at,
+                    immutable=asset_retention_reason is not None,
+                    retention_reason=asset_retention_reason,
+                    source_path=(
+                        f"/api/v1/files/{version.file_asset_id}/versions/"
+                        f"{version.id}/download"
+                    ),
+                )
+            )
+
+        for folder in _bounded_rows(
+            db.query(FileFolder)
+            .filter(
+                FileFolder.tenant_id == tenant_id,
+                or_(
+                    FileFolder.owner_user_id == subject_user_id,
+                    FileFolder.created_by_user_id == subject_user_id,
+                ),
+            )
+            .order_by(FileFolder.id)
+        ):
+            append(
+                _record(
+                    "file_folder",
+                    folder.id,
+                    "managed_folder",
+                    folder.path,
+                    {
+                        "match_fields": _matching_fields(
+                            folder,
+                            subject_user_id,
+                            ("owner_user_id", "created_by_user_id"),
+                        ),
+                        "owner_type": folder.owner_type,
+                        "path": folder.path,
+                        "deleted_at": _iso(folder.deleted_at),
+                    },
+                    observed_at=folder.updated_at,
+                    source_path="/files",
+                )
+            )
+
+        for share in _bounded_rows(
+            db.query(FileShare)
+            .filter(
+                FileShare.tenant_id == tenant_id,
+                or_(
+                    (FileShare.target_type == "user")
+                    & (FileShare.target_id == subject_user_id),
+                    FileShare.created_by_user_id == subject_user_id,
+                    FileShare.revoked_by_user_id == subject_user_id,
+                ),
+            )
+            .order_by(FileShare.id)
+        ):
+            target_matches = (
+                share.target_type == "user" and share.target_id == subject_user_id
+            )
+            match_fields = _matching_fields(
+                share,
+                subject_user_id,
+                ("created_by_user_id", "revoked_by_user_id"),
+            )
+            if target_matches:
+                match_fields.insert(0, "target_id")
+            append(
+                _record(
+                    "file_share",
+                    share.id,
+                    "file_access_evidence",
+                    f"Share for file {share.file_asset_id}",
+                    {
+                        "match_fields": match_fields,
+                        "file_asset_id": share.file_asset_id,
+                        "target_type": share.target_type,
+                        "target_is_subject": target_matches,
+                        "permission": share.permission,
+                        "expires_at": _iso(share.expires_at),
+                        "revoked_at": _iso(share.revoked_at),
+                    },
+                    observed_at=share.updated_at,
+                    immutable=True,
+                    retention_reason=(
+                        "File-sharing history is institutional access evidence."
+                    ),
+                    source_path=f"/files?file={share.file_asset_id}",
+                )
+            )
+
+        if form_evidence_available:
+            evidence_rows = _bounded_rows(
+                db.query(FileFormEvidenceGrant)
+                .outerjoin(
+                    FileAsset,
+                    FileAsset.id == FileFormEvidenceGrant.file_asset_id,
+                )
+                .filter(
+                    FileFormEvidenceGrant.tenant_id == tenant_id,
+                    or_(
+                        FileFormEvidenceGrant.custodian_user_id == subject_user_id,
+                        FileAsset.owner_user_id == subject_user_id,
+                    ),
+                )
+                .order_by(FileFormEvidenceGrant.id)
+            )
+            for grant in evidence_rows:
+                append(
+                    _record(
+                        "file_form_evidence",
+                        grant.id,
+                        "form_evidence",
+                        f"Form evidence {grant.form_definition_id}",
+                        {
+                            "match_fields": (
+                                ["custodian_user_id"]
+                                if grant.custodian_user_id == subject_user_id
+                                else ["asset.owner_user_id"]
+                            ),
+                            "form_instance_id": grant.form_instance_id,
+                            "form_definition_id": grant.form_definition_id,
+                            "form_definition_revision": grant.form_definition_revision,
+                            "evidence_kind": grant.evidence_kind,
+                            "purpose": grant.purpose,
+                            "status": grant.status,
+                            "expires_at": _iso(grant.expires_at),
+                            "file_asset_id": grant.file_asset_id,
+                            "file_version_id": grant.file_version_id,
+                        },
+                        observed_at=grant.updated_at,
+                        immutable=True,
+                        retention_reason=(
+                            "Submitted Form attachment evidence follows the owning "
+                            "process retention and cannot be erased through Files alone."
+                        ),
+                    )
+                )
+
+        if campaign_evidence_available:
+            campaign_rows = _bounded_rows(
+                db.query(CampaignAttachmentUse)
+                .join(FileAsset, FileAsset.id == CampaignAttachmentUse.file_asset_id)
+                .filter(
+                    CampaignAttachmentUse.tenant_id == tenant_id,
+                    FileAsset.owner_user_id == subject_user_id,
+                )
+                .order_by(CampaignAttachmentUse.id)
+            )
+            for use in campaign_rows:
+                append(
+                    _record(
+                        "campaign_attachment_use",
+                        use.id,
+                        "delivery_evidence",
+                        use.filename_used,
+                        {
+                            "match_fields": ["asset.owner_user_id"],
+                            "campaign_id": use.campaign_id,
+                            "campaign_version_id": use.campaign_version_id,
+                            "campaign_job_id": use.campaign_job_id,
+                            "file_asset_id": use.file_asset_id,
+                            "file_version_id": use.file_version_id,
+                            "filename_used": use.filename_used,
+                            "checksum_sha256": use.checksum_sha256,
+                            "size_bytes": use.size_bytes,
+                            "use_stage": use.use_stage,
+                            "used_at": _iso(use.used_at),
+                        },
+                        observed_at=use.updated_at,
+                        immutable=True,
+                        retention_reason=(
+                            "Campaign attachment use is immutable delivery evidence."
+                        ),
+                    )
+                )
+
+        self._append_configuration_references(
+            db,
+            append=append,
+            tenant_id=tenant_id,
+            subject_user_id=subject_user_id,
+        )
+        self._append_integrity_references(
+            db,
+            append=append,
+            tenant_id=tenant_id,
+            subject_user_id=subject_user_id,
+        )
+        return tuple(records)
+
+    def plan_erasure(
+        self,
+        session: object,
+        *,
+        tenant_id: str,
+        subject: DsarSubjectRef,
+        records: Sequence[DsarRecordRef],
+    ) -> Sequence[DsarErasureActionRef]:
+        del session
+        subject_user_id = _subject_user_id(subject)
+        if subject_user_id is None:
+            return ()
+        actions: list[DsarErasureActionRef] = []
+        for record in records:
+            if record.provider_id != self.provider_id or record.module_id != self.module_id:
+                raise ValueError("Files DSAR received a foreign provider record.")
+            match_fields = {
+                str(value) for value in record.data.get("match_fields", ())
+            }
+            if record.immutable_evidence:
+                actions.append(
+                    _action(
+                        f"files:retain:{record.resource_type}:{record.resource_id}",
+                        "retain",
+                        record,
+                        f"Retain {record.title}",
+                        record.retention_reason
+                        or "Institutional evidence must be retained.",
+                        executable=False,
+                    )
+                )
+            elif record.resource_type in {
+                "file_asset",
+                "file_version",
+                "file_folder",
+            }:
+                actions.append(
+                    _action(
+                        f"files:review:{record.resource_type}:{record.resource_id}",
+                        "manual_review",
+                        record,
+                        f"Review {record.title}",
+                        (
+                            "Managed file content, names, paths, ownership, and shared "
+                            "references require a case decision. Approved byte erasure "
+                            "must use the separate Files purge workflow."
+                        ),
+                        executable=False,
+                    )
+                )
+            if (
+                record.resource_type == "file_share"
+                and "target_id" in match_fields
+                and record.data.get("revoked_at") is None
+            ):
+                actions.append(
+                    _action(
+                        f"files:revoke:file_share:{record.resource_id}",
+                        "revoke",
+                        record,
+                        "Revoke active file share",
+                        "The active user-targeted share can be revoked without deleting file evidence.",
+                        executable=True,
+                        metadata={"subject_user_id": subject_user_id},
+                    )
+                )
+            for field_name in sorted(
+                match_fields.intersection(_detachable_fields(record.resource_type))
+            ):
+                actions.append(
+                    _action(
+                        (
+                            f"files:detach:{record.resource_type}:"
+                            f"{field_name}:{record.resource_id}"
+                        ),
+                        "detach",
+                        record,
+                        f"Detach {field_name.replace('_', ' ')}",
+                        (
+                            "Remove the mutable subject reference while preserving the "
+                            "governed resource and DSAR evidence."
+                        ),
+                        executable=True,
+                        metadata={
+                            "field": field_name,
+                            "subject_user_id": subject_user_id,
+                        },
+                    )
+                )
+        action_ids = [action.action_id for action in actions]
+        if len(action_ids) != len(set(action_ids)):
+            raise ValueError("Files DSAR produced duplicate action ids.")
+        return tuple(actions)
+
+    def execute_erasure(
+        self,
+        session: object,
+        *,
+        tenant_id: str,
+        subject: DsarSubjectRef,
+        actions: Sequence[DsarErasureActionRef],
+        request_id: str,
+    ) -> Sequence[DsarExecutionResultRef]:
+        db = _session(session)
+        subject_user_id = _subject_user_id(subject)
+        if subject_user_id is None:
+            return tuple(
+                _blocked(action, "Files requires a direct membership subject reference.")
+                for action in actions
+            )
+        results: list[DsarExecutionResultRef] = []
+        for action in actions:
+            if (
+                action.provider_id != self.provider_id
+                or action.module_id != self.module_id
+                or action.metadata.get("subject_user_id") != subject_user_id
+            ):
+                results.append(_blocked(action, "The Files DSAR action is stale or invalid."))
+                continue
+            if action.action_id.startswith("files:revoke:file_share:"):
+                results.append(
+                    _revoke_share(
+                        db,
+                        tenant_id=tenant_id,
+                        subject_user_id=subject_user_id,
+                        action=action,
+                        request_id=request_id,
+                    )
+                )
+            elif action.action_id.startswith("files:detach:"):
+                results.append(
+                    _detach_reference(
+                        db,
+                        tenant_id=tenant_id,
+                        subject_user_id=subject_user_id,
+                        action=action,
+                        request_id=request_id,
+                    )
+                )
+            else:
+                results.append(_blocked(action, "Files does not execute this action kind."))
+        db.flush()
+        return tuple(results)
+
+    def _append_configuration_references(
+        self,
+        db: Session,
+        *,
+        append: object,
+        tenant_id: str,
+        subject_user_id: str,
+    ) -> None:
+        configurations = (
+            (FileConnectorProfile, "connector_profile", ("created_by_user_id", "updated_by_user_id")),
+            (FileConnectorCredential, "connector_credential", ("created_by_user_id", "updated_by_user_id")),
+            (FileConnectorPolicy, "connector_policy", ("created_by_user_id", "updated_by_user_id")),
+            (FileConnectorSpace, "connector_space", ("owner_user_id", "created_by_user_id")),
+        )
+        for model, resource_type, fields in configurations:
+            if not _has_table(db, model):
+                continue
+            conditions = [getattr(model, field) == subject_user_id for field in fields]
+            query = db.query(model).filter(or_(*conditions))
+            if hasattr(model, "tenant_id"):
+                query = query.filter(model.tenant_id == tenant_id)
+            for row in _bounded_rows(query.order_by(model.id)):
+                match_fields = _matching_fields(row, subject_user_id, fields)
+                data: dict[str, object] = {
+                    "match_fields": match_fields,
+                    "label": getattr(row, "label", None),
+                    "provider": getattr(row, "provider", None),
+                }
+                if isinstance(row, FileConnectorCredential):
+                    data["credential_mode"] = row.credential_mode
+                elif isinstance(row, FileConnectorSpace):
+                    data.update(
+                        {
+                            "remote_path": row.remote_path,
+                            "sync_mode": row.sync_mode,
+                            "read_only": row.read_only,
+                            "deleted_at": _iso(row.deleted_at),
+                        }
+                    )
+                append(  # type: ignore[operator]
+                    _record(
+                        resource_type,
+                        row.id,
+                        "connector_configuration_evidence",
+                        getattr(row, "label", None) or resource_type.replace("_", " "),
+                        data,
+                        observed_at=row.updated_at,
+                        immutable=True,
+                        retention_reason=(
+                            "Connector configuration history is institutional evidence; "
+                            "credential secrets are excluded from the DSAR export."
+                        ),
+                    )
+                )
+
+    def _append_integrity_references(
+        self,
+        db: Session,
+        *,
+        append: object,
+        tenant_id: str,
+        subject_user_id: str,
+    ) -> None:
+        if _has_table(db, FileIntegrityScan):
+            for scan in _bounded_rows(
+                db.query(FileIntegrityScan)
+                .filter(
+                    FileIntegrityScan.tenant_id == tenant_id,
+                    FileIntegrityScan.created_by_user_id == subject_user_id,
+                )
+                .order_by(FileIntegrityScan.id)
+            ):
+                append(  # type: ignore[operator]
+                    _record(
+                        "file_integrity_scan",
+                        scan.id,
+                        "storage_integrity_evidence",
+                        f"Integrity scan {scan.id}",
+                        {
+                            "match_fields": ["created_by_user_id"],
+                            "storage_backend": scan.storage_backend,
+                            "status": scan.status,
+                            "started_at": _iso(scan.started_at),
+                            "completed_at": _iso(scan.completed_at),
+                        },
+                        observed_at=scan.updated_at,
+                        immutable=True,
+                        retention_reason="Storage integrity scans are operator evidence.",
+                    )
+                )
+        if _has_table(db, FileIntegrityFinding):
+            for finding in _bounded_rows(
+                db.query(FileIntegrityFinding)
+                .filter(
+                    FileIntegrityFinding.tenant_id == tenant_id,
+                    FileIntegrityFinding.resolved_by_user_id == subject_user_id,
+                )
+                .order_by(FileIntegrityFinding.id)
+            ):
+                append(  # type: ignore[operator]
+                    _record(
+                        "file_integrity_finding",
+                        finding.id,
+                        "storage_integrity_evidence",
+                        f"Integrity finding {finding.kind}",
+                        {
+                            "match_fields": ["resolved_by_user_id"],
+                            "kind": finding.kind,
+                            "state": finding.state,
+                            "resolved_at": _iso(finding.resolved_at),
+                        },
+                        observed_at=finding.updated_at,
+                        immutable=True,
+                        retention_reason="Integrity resolution is operator evidence.",
+                    )
+                )
+
+
+def _asset_retention_reason(
+    session: Session,
+    asset: FileAsset,
+    *,
+    form_evidence_available: bool,
+    campaign_evidence_available: bool,
+) -> str | None:
+    reasons: list[str] = []
+    if asset.legal_hold:
+        reasons.append("The file is under legal hold.")
+    retained_until = _aware(asset.retained_until)
+    if retained_until is not None and retained_until > datetime.now(timezone.utc):
+        reasons.append(f"The file is retained until {retained_until.isoformat()}.")
+    if form_evidence_available:
+        has_form_evidence = (
+            session.query(FileFormEvidenceGrant.id)
+            .filter(
+                FileFormEvidenceGrant.tenant_id == asset.tenant_id,
+                FileFormEvidenceGrant.file_asset_id == asset.id,
+            )
+            .first()
+            is not None
+        )
+        if has_form_evidence:
+            reasons.append("The file is referenced by submitted Form evidence.")
+    if campaign_evidence_available:
+        has_campaign_evidence = (
+            session.query(CampaignAttachmentUse.id)
+            .filter(
+                CampaignAttachmentUse.tenant_id == asset.tenant_id,
+                CampaignAttachmentUse.file_asset_id == asset.id,
+            )
+            .first()
+            is not None
+        )
+        if has_campaign_evidence:
+            reasons.append("The file is referenced by Campaign delivery evidence.")
+    return " ".join(reasons) or None
+
+
+_DETACHABLE_MODELS: dict[str, tuple[type[object], frozenset[str]]] = {
+    "file_asset": (FileAsset, frozenset({"created_by_user_id"})),
+    "file_version": (FileVersion, frozenset({"created_by_user_id"})),
+    "file_folder": (FileFolder, frozenset({"created_by_user_id"})),
+    "file_share": (
+        FileShare,
+        frozenset({"created_by_user_id", "revoked_by_user_id"}),
+    ),
+    "connector_profile": (
+        FileConnectorProfile,
+        frozenset({"created_by_user_id", "updated_by_user_id"}),
+    ),
+    "connector_credential": (
+        FileConnectorCredential,
+        frozenset({"created_by_user_id", "updated_by_user_id"}),
+    ),
+    "connector_policy": (
+        FileConnectorPolicy,
+        frozenset({"created_by_user_id", "updated_by_user_id"}),
+    ),
+    "connector_space": (FileConnectorSpace, frozenset({"created_by_user_id"})),
+}
+
+
+def _detachable_fields(resource_type: str) -> frozenset[str]:
+    entry = _DETACHABLE_MODELS.get(resource_type)
+    return entry[1] if entry is not None else frozenset()
+
+
+def _revoke_share(
+    session: Session,
+    *,
+    tenant_id: str,
+    subject_user_id: str,
+    action: DsarErasureActionRef,
+    request_id: str,
+) -> DsarExecutionResultRef:
+    row = (
+        session.query(FileShare)
+        .filter(FileShare.id == action.resource_id)
+        .with_for_update()
+        .one_or_none()
+    )
+    if (
+        row is None
+        or row.tenant_id != tenant_id
+        or row.target_type != "user"
+        or row.target_id != subject_user_id
+    ):
+        return _blocked(action, "The subject-targeted file share is no longer available.")
+    if row.revoked_at is not None:
+        return _result(
+            action,
+            "unchanged",
+            "The file share was already revoked.",
+            {"request_id": request_id, "revoked_at": _iso(row.revoked_at)},
+        )
+    row.revoked_at = datetime.now(timezone.utc)
+    row.revoked_by_user_id = None
+    return _result(
+        action,
+        "executed",
+        "The subject-targeted file share was revoked.",
+        {"request_id": request_id, "revoked_at": _iso(row.revoked_at)},
+    )
+
+
+def _detach_reference(
+    session: Session,
+    *,
+    tenant_id: str,
+    subject_user_id: str,
+    action: DsarErasureActionRef,
+    request_id: str,
+) -> DsarExecutionResultRef:
+    entry = _DETACHABLE_MODELS.get(action.resource_type)
+    field_name = str(action.metadata.get("field") or "")
+    if entry is None or field_name not in entry[1]:
+        return _blocked(action, "The requested Files subject reference is not detachable.")
+    model = entry[0]
+    row = (
+        session.query(model)
+        .filter(getattr(model, "id") == action.resource_id)
+        .with_for_update()
+        .one_or_none()
+    )
+    if row is None or getattr(row, "tenant_id", None) != tenant_id:
+        return _blocked(action, "The Files resource is no longer available.")
+    current = getattr(row, field_name)
+    if current is None:
+        return _result(
+            action,
+            "unchanged",
+            "The subject reference was already detached.",
+            {"request_id": request_id, "field": field_name},
+        )
+    if current != subject_user_id:
+        return _blocked(action, "The Files subject reference changed after planning.")
+    setattr(row, field_name, None)
+    return _result(
+        action,
+        "executed",
+        "The mutable subject reference was detached.",
+        {"request_id": request_id, "field": field_name},
+    )
+
+
+def _subject_user_id(subject: DsarSubjectRef) -> str | None:
+    candidates: list[str] = []
+    if subject.membership_id:
+        candidates.append(subject.membership_id)
+    for key, value in subject.external_references.items():
+        if key in {
+            "files.user",
+            "files.membership",
+            "access.membership",
+            "membership_id",
+        }:
+            candidates.append(value)
+    normalized = {value.strip() for value in candidates if value.strip()}
+    if len(normalized) != 1:
+        return None
+    return normalized.pop()
+
+
+def _matching_fields(
+    row: object,
+    subject_user_id: str,
+    fields: Sequence[str],
+) -> list[str]:
+    return [field for field in fields if getattr(row, field) == subject_user_id]
+
+
+def _record(
+    resource_type: str,
+    resource_id: str,
+    category: str,
+    title: str,
+    data: dict[str, object],
+    *,
+    observed_at: datetime | None = None,
+    immutable: bool = False,
+    retention_reason: str | None = None,
+    source_path: str | None = None,
+) -> DsarRecordRef:
+    return DsarRecordRef(
+        provider_id="files",
+        module_id="files",
+        resource_type=resource_type,
+        resource_id=resource_id,
+        category=category,
+        title=title,
+        data=data,
+        observed_at=observed_at,
+        immutable_evidence=immutable,
+        retention_reason=retention_reason,
+        source_path=source_path,
+    )
+
+
+def _action(
+    action_id: str,
+    kind: str,
+    record: DsarRecordRef,
+    title: str,
+    rationale: str,
+    *,
+    executable: bool,
+    metadata: dict[str, object] | None = None,
+) -> DsarErasureActionRef:
+    return DsarErasureActionRef(
+        action_id=action_id,
+        provider_id="files",
+        module_id="files",
+        kind=kind,  # type: ignore[arg-type]
+        resource_type=record.resource_type,
+        resource_id=record.resource_id,
+        title=title,
+        rationale=rationale,
+        executable=executable,
+        metadata=metadata or {},
+    )
+
+
+def _result(
+    action: DsarErasureActionRef,
+    status: str,
+    summary: str,
+    evidence: dict[str, object] | None = None,
+) -> DsarExecutionResultRef:
+    return DsarExecutionResultRef(
+        action_id=action.action_id,
+        status=status,  # type: ignore[arg-type]
+        summary=summary,
+        evidence=evidence or {},
+    )
+
+
+def _blocked(action: DsarErasureActionRef, summary: str) -> DsarExecutionResultRef:
+    return _result(action, "blocked", summary)
+
+
+def _session(value: object) -> Session:
+    if not isinstance(value, Session):
+        raise TypeError("Files DSAR provider requires a SQLAlchemy session.")
+    return value
+
+
+def _bounded_rows(query: object) -> list[object]:
+    rows = query.limit(_MAX_RECORDS + 1).all()  # type: ignore[attr-defined]
+    if len(rows) > _MAX_RECORDS:
+        raise ValueError("Files DSAR match limit exceeded; narrow the subject selectors.")
+    return rows
+
+
+def _has_table(session: Session, model: type[object]) -> bool:
+    return inspect(session.connection()).has_table(model.__tablename__)
+
+
+def _aware(value: datetime | None) -> datetime | None:
+    if value is None or value.tzinfo is not None:
+        return value
+    return value.replace(tzinfo=timezone.utc)
+
+
+def _iso(value: datetime | None) -> str | None:
+    aware = _aware(value)
+    return aware.isoformat() if aware else None
+
+
+__all__ = ["FILES_DSAR_CAPABILITY", "FilesDsarProvider"]

src/govoplan_files/backend/manifest.py

@@ -54,6 +54,7 @@ from govoplan_files.backend.configuration_provider import (
 )
 from govoplan_files.backend.db import models as file_models  # noqa: F401 - populate Files ORM metadata
 from govoplan_files.backend.documentation import documentation_topics
+from govoplan_files.backend.dsar_provider import FILES_DSAR_CAPABILITY
 from govoplan_files.backend.form_evidence import (
     CAPABILITY_FORM_EVIDENCE_FILES,
     create_files_form_evidence_provider,
@@ -440,6 +441,13 @@ REMOTE_STORAGE_PROVIDER = ExternalProviderDeclaration(
 )
 
 
+def _dsar_provider(context: ModuleContext) -> object:
+    del context
+    from govoplan_files.backend.dsar_provider import FilesDsarProvider
+
+    return FilesDsarProvider()
+
+
 manifest = ModuleManifest(
     id="files",
     name="Files",
@@ -456,6 +464,7 @@ manifest = ModuleManifest(
         ModuleInterfaceProvider(name=CAPABILITY_FILES_POSTBOX_REFERENCES, version="1.0.0"),
         ModuleInterfaceProvider(name=CAPABILITY_RECORD_SOURCE_FILES, version="1.0.0"),
         ModuleInterfaceProvider(name=CAPABILITY_FORM_EVIDENCE_FILES, version="1.0.0"),
+        ModuleInterfaceProvider(name=FILES_DSAR_CAPABILITY, version="0.1.0"),
     ),
     requires_interfaces=(
         ModuleInterfaceRequirement(
@@ -1180,6 +1189,69 @@ manifest = ModuleManifest(
                 ],
             },
         ),
+        DocumentationTopic(
+            id="files.privacy.data-subject-requests",
+            title="Review Files data in a data-subject request",
+            summary="Collect safe Files metadata and keep retention, evidence, and byte-erasure decisions explicit.",
+            body=(
+                "The Files DSAR provider searches only the effective tenant and requires a direct membership or namespaced Files user reference. "
+                "It exports bounded file, version, folder, sharing, evidence, connector-configuration, and integrity metadata without raw file bytes, storage locations, tokens, passwords, secret references, or encrypted credential values. "
+                "Plans may revoke an active share aimed at the subject or detach a mutable actor reference. Legal hold, active retention, Form evidence, Campaign delivery evidence, configuration history, and integrity evidence remain retained with a reason. File content, ownership, names, and paths require manual review. Approved physical erasure must use the separately authorized Files purge and blob-garbage-collection workflow so DSAR execution cannot bypass evidence blockers, approval, audit, or recovery controls."
+            ),
+            layer="configured",
+            documentation_types=("admin",),
+            audience=("privacy_officer", "file_admin", "records_manager", "operator"),
+            order=47,
+            conditions=(
+                DocumentationCondition(
+                    required_modules=("files", "access"),
+                    any_scopes=(
+                        "access:privacy:read",
+                        "access:privacy:manage",
+                        "access:privacy:erase",
+                    ),
+                ),
+            ),
+            links=(
+                DocumentationLink(
+                    label="Data-subject requests",
+                    href="/admin?section=tenant-data-subject-requests",
+                    kind="runtime",
+                ),
+                DocumentationLink(
+                    label="Files handbook",
+                    href="govoplan-files/docs/FILES_HANDBOOK.md",
+                    kind="repository",
+                ),
+            ),
+            related_modules=("access", "audit", "campaigns", "forms-runtime", "ops"),
+            metadata={
+                "kind": "workflow",
+                "route": "/admin?section=tenant-data-subject-requests",
+                "screen": "Data-subject requests",
+                "help_contexts": ["admin.privacy.data-subject-requests"],
+                "prerequisites": [
+                    "The request has been authorized and contains a direct tenant membership or Files subject reference.",
+                    "The privacy reviewer can distinguish access export from erasure authority and Files purge authority.",
+                ],
+                "steps": [
+                    "Run the provider search and confirm Files reports complete coverage rather than a failed or absent provider.",
+                    "Review file/version metadata, evidence retention reasons, and the source path for manual content review.",
+                    "Generate the erasure plan and execute only the approved reversible share-revocation or subject-reference actions.",
+                    "For approved byte erasure, resolve every lifecycle blocker and use Files purge preview, execution, and blob garbage collection separately.",
+                ],
+                "limitations": [
+                    "Email, account, or identity selectors alone cannot be resolved by Files because Files does not own the Access directory; supply the corroborated membership reference.",
+                    "The provider does not embed raw file content in the JSON export and never performs physical blob deletion as a DSAR side effect.",
+                ],
+                "outcome": "Files-owned subject references are reviewed or removed without silently destroying retained content or evidence.",
+                "verification": "Confirm every Files record has a retain, review, revoke, or detach disposition and inspect any separate purge through its audit and recovery evidence.",
+                "related_topic_ids": [
+                    "files.workflow.restore-retain-and-purge",
+                    "files.reference.integrity-recovery-and-fail-closed-transports",
+                ],
+            },
+        ),
         DocumentationTopic(
             id="files.governed-connectors-and-provenance",
             title="Govern file connections and credential deletion",
@@ -1635,6 +1707,7 @@ manifest = ModuleManifest(
         ).campaign_capability(context),
         CAPABILITY_RECORD_SOURCE_FILES: create_files_record_source,
         CAPABILITY_FORM_EVIDENCE_FILES: create_files_form_evidence_provider,
+        FILES_DSAR_CAPABILITY: _dsar_provider,
     },
     capability_documentation={
         CAPABILITY_RECORD_SOURCE_FILES: CapabilityDocumentation(
@@ -1647,6 +1720,11 @@ manifest = ModuleManifest(
             summary="Issues one-time managed attachment grants and verifies exact Form evidence versions.",
             contract_version="1.0.0",
         ),
+        FILES_DSAR_CAPABILITY: CapabilityDocumentation(
+            label="Files data-subject request provider",
+            summary="Finds safe Files metadata and classifies reversible references, manual file review, and retained evidence.",
+            contract_version="0.1.0",
+        ),
     },
     operational_check_providers=(
         OperationalCheckProviderRegistration(

tests/test_dsar_provider.py

@@ -0,0 +1,468 @@
+from __future__ import annotations
+
+import unittest
+from datetime import datetime, timedelta, timezone
+
+from sqlalchemy import create_engine
+from sqlalchemy.orm import sessionmaker
+
+from govoplan_access.backend.db.models import Account, Group, User
+from govoplan_core.core.dsar import DsarProvider, DsarSubjectRef
+from govoplan_core.core.change_sequence import ChangeSequenceEntry
+from govoplan_core.db.base import Base
+from govoplan_core.privacy.dsar_workflow import (
+    DataSubjectRequest,
+    create_data_subject_request,
+    execute_data_subject_erasure,
+    plan_data_subject_erasure,
+    search_data_subject_request,
+)
+from govoplan_files.backend.db.models import (
+    FileAsset,
+    FileBlob,
+    FileConnectorCredential,
+    FileConnectorPolicy,
+    FileConnectorProfile,
+    FileConnectorSpace,
+    FileFolder,
+    FileFormEvidenceGrant,
+    FileIntegrityFinding,
+    FileIntegrityScan,
+    FileShare,
+    FileVersion,
+)
+from govoplan_files.backend.dsar_provider import (
+    FILES_DSAR_CAPABILITY,
+    FilesDsarProvider,
+)
+from govoplan_files.backend.manifest import manifest
+
+
+class _Registry:
+    def __init__(self, provider: FilesDsarProvider, *, files_active: bool = True) -> None:
+        self.provider = provider
+        self.files_active = files_active
+
+    def capability_names(self):
+        return (FILES_DSAR_CAPABILITY,)
+
+    def capability_owner(self, name):
+        self._assert_capability(name)
+        return "files"
+
+    def tenant_entitlement_resolver(self):
+        files_active = self.files_active
+
+        class _Resolver:
+            @staticmethod
+            def resolve(session, tenant_id):
+                del session, tenant_id
+                return type(
+                    "State",
+                    (),
+                    {"effective_modules": ("files",) if files_active else ()},
+                )()
+
+        return _Resolver()
+
+    def require_tenant_capability(self, name, session, **kwargs):
+        del session, kwargs
+        self._assert_capability(name)
+        return self.provider
+
+    def manifests(self):
+        return (type("Manifest", (), {"id": "files"})(),)
+
+    @staticmethod
+    def _assert_capability(name: str) -> None:
+        if name != FILES_DSAR_CAPABILITY:
+            raise KeyError(name)
+
+
+class FilesDsarProviderTests(unittest.TestCase):
+    def setUp(self) -> None:
+        self.engine = create_engine("sqlite:///:memory:", future=True)
+        Base.metadata.create_all(
+            bind=self.engine,
+            tables=[
+                Account.__table__,
+                User.__table__,
+                Group.__table__,
+                ChangeSequenceEntry.__table__,
+                DataSubjectRequest.__table__,
+                FileBlob.__table__,
+                FileAsset.__table__,
+                FileVersion.__table__,
+                FileFolder.__table__,
+                FileFormEvidenceGrant.__table__,
+                FileShare.__table__,
+                FileConnectorCredential.__table__,
+                FileConnectorPolicy.__table__,
+                FileConnectorProfile.__table__,
+                FileConnectorSpace.__table__,
+                FileIntegrityScan.__table__,
+                FileIntegrityFinding.__table__,
+            ],
+        )
+        self.session = sessionmaker(bind=self.engine, future=True)()
+        self.account = Account(
+            id="account-1",
+            email="subject@example.test",
+            normalized_email="subject@example.test",
+            display_name="Subject",
+            password_hash="not-exported",
+        )
+        self.user = User(
+            id="membership-1",
+            tenant_id="tenant-1",
+            account_id=self.account.id,
+            email="subject@example.test",
+            display_name="Subject",
+        )
+        self.blob = FileBlob(
+            id="blob-1",
+            tenant_id="tenant-1",
+            storage_backend="local",
+            storage_key="private/storage/key-do-not-export",
+            checksum_sha256="a" * 64,
+            size_bytes=12,
+            ref_count=1,
+        )
+        self.asset = FileAsset(
+            id="asset-1",
+            tenant_id="tenant-1",
+            owner_type="user",
+            owner_user_id=self.user.id,
+            created_by_user_id=self.user.id,
+            current_version_id="version-1",
+            display_path="subjects/private.txt",
+            filename="private.txt",
+            description="Subject-provided document",
+            retained_until=datetime.now(timezone.utc) + timedelta(days=30),
+            lifecycle_reason="Pending proceeding",
+            metadata_={"password": "metadata-secret-do-not-export"},
+        )
+        self.version = FileVersion(
+            id="version-1",
+            tenant_id="tenant-1",
+            file_asset_id=self.asset.id,
+            blob_id=self.blob.id,
+            version_number=1,
+            filename_at_upload=self.asset.filename,
+            display_path_at_upload=self.asset.display_path,
+            content_type="text/plain",
+            size_bytes=12,
+            checksum_sha256="a" * 64,
+            created_by_user_id=self.user.id,
+        )
+        self.folder = FileFolder(
+            id="folder-1",
+            tenant_id="tenant-1",
+            owner_type="user",
+            owner_user_id=self.user.id,
+            path="subjects",
+            created_by_user_id=self.user.id,
+        )
+        self.share = FileShare(
+            id="share-1",
+            tenant_id="tenant-1",
+            file_asset_id=self.asset.id,
+            target_type="user",
+            target_id=self.user.id,
+            permission="read",
+            created_by_user_id=self.user.id,
+        )
+        self.form_evidence = FileFormEvidenceGrant(
+            id="evidence-1",
+            tenant_id="tenant-1",
+            form_instance_id="form-instance-1",
+            form_definition_id="application",
+            form_definition_revision="7",
+            token_sha256="b" * 64,
+            idempotency_key="evidence-key-1",
+            request_sha256="c" * 64,
+            custodian_user_id=self.user.id,
+            evidence_kind="attachment",
+            purpose="Submitted application evidence",
+            status="uploaded",
+            expires_at=datetime.now(timezone.utc) + timedelta(days=1),
+            max_size_bytes=1024,
+            allowed_content_types=["text/plain"],
+            file_asset_id=self.asset.id,
+            file_version_id=self.version.id,
+            metadata_={},
+        )
+        self.credential = FileConnectorCredential(
+            id="credential-1",
+            tenant_id="tenant-1",
+            scope_type="tenant",
+            scope_id="tenant-1",
+            label="Subject-created credential",
+            provider="s3",
+            credential_mode="database",
+            username="subject-user",
+            password_encrypted="encrypted-password-do-not-export",
+            token_encrypted="encrypted-token-do-not-export",
+            password_env="PASSWORD_ENV_DO_NOT_EXPORT",
+            secret_ref="vault://do-not-export",
+            created_by_user_id=self.user.id,
+            updated_by_user_id=self.user.id,
+        )
+        self.integrity_scan = FileIntegrityScan(
+            id="scan-1",
+            tenant_id="tenant-1",
+            storage_backend="local",
+            storage_prefix="private-prefix-do-not-export",
+            created_by_user_id=self.user.id,
+        )
+        tenant_two_asset = FileAsset(
+            id="asset-tenant-2",
+            tenant_id="tenant-2",
+            owner_type="user",
+            owner_user_id=self.user.id,
+            display_path="other-tenant.txt",
+            filename="other-tenant.txt",
+        )
+        self.session.add_all(
+            [
+                self.account,
+                self.user,
+                self.blob,
+                self.asset,
+                self.version,
+                self.folder,
+                self.share,
+                self.form_evidence,
+                self.credential,
+                self.integrity_scan,
+                tenant_two_asset,
+            ]
+        )
+        self.session.commit()
+        self.provider = FilesDsarProvider()
+        self.subject = DsarSubjectRef(membership_id=self.user.id)
+
+    def tearDown(self) -> None:
+        self.session.close()
+        self.engine.dispose()
+
+    def test_manifest_publishes_protocol_conforming_provider(self) -> None:
+        provided_names = {item.name for item in manifest.provides_interfaces}
+        self.assertIn(FILES_DSAR_CAPABILITY, provided_names)
+        provider = manifest.capability_factories[FILES_DSAR_CAPABILITY](None)
+        self.assertIsInstance(provider, DsarProvider)
+
+    def test_search_is_tenant_scoped_and_excludes_bytes_and_secrets(self) -> None:
+        records = self.provider.search_subject(
+            self.session,
+            tenant_id="tenant-1",
+            subject=self.subject,
+        )
+        resource_types = {record.resource_type for record in records}
+        self.assertTrue(
+            {
+                "file_asset",
+                "file_version",
+                "file_folder",
+                "file_share",
+                "file_form_evidence",
+                "connector_credential",
+                "file_integrity_scan",
+            }.issubset(resource_types)
+        )
+        serialized = repr([record.to_dict() for record in records])
+        self.assertNotIn("asset-tenant-2", serialized)
+        self.assertNotIn("private/storage/key-do-not-export", serialized)
+        self.assertNotIn("metadata-secret-do-not-export", serialized)
+        self.assertNotIn("encrypted-password-do-not-export", serialized)
+        self.assertNotIn("encrypted-token-do-not-export", serialized)
+        self.assertNotIn("subject-user", serialized)
+        self.assertNotIn("PASSWORD_ENV_DO_NOT_EXPORT", serialized)
+        self.assertNotIn("vault://do-not-export", serialized)
+        self.assertNotIn("private-prefix-do-not-export", serialized)
+
+    def test_conflicting_direct_subject_references_fail_closed(self) -> None:
+        records = self.provider.search_subject(
+            self.session,
+            tenant_id="tenant-1",
+            subject=DsarSubjectRef(
+                membership_id=self.user.id,
+                external_references={"files.user": "another-membership"},
+            ),
+        )
+
+        self.assertEqual((), records)
+
+    def test_plan_classifies_retention_and_manual_review(self) -> None:
+        records = self.provider.search_subject(
+            self.session,
+            tenant_id="tenant-1",
+            subject=self.subject,
+        )
+        actions = self.provider.plan_erasure(
+            self.session,
+            tenant_id="tenant-1",
+            subject=self.subject,
+            records=records,
+        )
+
+        kinds = {action.kind for action in actions}
+        self.assertTrue({"retain", "manual_review", "revoke", "detach"}.issubset(kinds))
+        asset_retention = next(
+            action
+            for action in actions
+            if action.action_id == "files:retain:file_asset:asset-1"
+        )
+        self.assertIn("retained until", asset_retention.rationale)
+        self.assertTrue(
+            any(action.action_id == "files:revoke:file_share:share-1" for action in actions)
+        )
+        self.assertFalse(
+            any(action.kind == "delete" and action.executable for action in actions)
+        )
+
+    def test_execution_is_revalidated_and_idempotent(self) -> None:
+        records = self.provider.search_subject(
+            self.session,
+            tenant_id="tenant-1",
+            subject=self.subject,
+        )
+        actions = self.provider.plan_erasure(
+            self.session,
+            tenant_id="tenant-1",
+            subject=self.subject,
+            records=records,
+        )
+        executable = tuple(action for action in actions if action.executable)
+
+        first = self.provider.execute_erasure(
+            self.session,
+            tenant_id="tenant-1",
+            subject=self.subject,
+            actions=executable,
+            request_id="dsar-1",
+        )
+        self.assertEqual({"executed"}, {result.status for result in first})
+        self.assertIsNotNone(self.share.revoked_at)
+        self.assertIsNone(self.asset.created_by_user_id)
+        self.assertIsNone(self.version.created_by_user_id)
+        self.assertIsNone(self.folder.created_by_user_id)
+        self.assertIsNone(self.credential.created_by_user_id)
+        self.assertIsNone(self.credential.updated_by_user_id)
+
+        repeated = self.provider.execute_erasure(
+            self.session,
+            tenant_id="tenant-1",
+            subject=self.subject,
+            actions=executable,
+            request_id="dsar-1",
+        )
+        self.assertEqual({"unchanged"}, {result.status for result in repeated})
+
+    def test_execution_blocks_when_reference_changed_after_planning(self) -> None:
+        records = self.provider.search_subject(
+            self.session,
+            tenant_id="tenant-1",
+            subject=self.subject,
+        )
+        actions = self.provider.plan_erasure(
+            self.session,
+            tenant_id="tenant-1",
+            subject=self.subject,
+            records=records,
+        )
+        detach = next(
+            action
+            for action in actions
+            if action.action_id
+            == "files:detach:file_version:created_by_user_id:version-1"
+        )
+        self.version.created_by_user_id = "replacement-user"
+        self.session.flush()
+
+        result = self.provider.execute_erasure(
+            self.session,
+            tenant_id="tenant-1",
+            subject=self.subject,
+            actions=(detach,),
+            request_id="dsar-2",
+        )
+
+        self.assertEqual("blocked", result[0].status)
+        self.assertEqual("replacement-user", self.version.created_by_user_id)
+
+    def test_core_workflow_discovers_active_provider_and_skips_it_when_disabled(
+        self,
+    ) -> None:
+        request = create_data_subject_request(
+            self.session,
+            tenant_id="tenant-1",
+            reference="DSAR-FILES-1",
+            request_kind="access_and_erasure",
+            subject=self.subject,
+            purpose="Respond to an authorized privacy request.",
+            legal_basis="Article 15 and 17 GDPR",
+            due_at=None,
+            requested_by_account_id="privacy-officer",
+        )
+        self.session.commit()
+        registry = _Registry(self.provider)
+
+        search_data_subject_request(
+            self.session,
+            registry=registry,
+            row=request,
+            expected_revision=1,
+        )
+        self.assertEqual(
+            "searched", request.status, request.search_result["provider_runs"]
+        )
+        self.assertEqual(["files"], request.coverage["covered_modules"])
+        self.assertEqual([], request.coverage["modules_without_provider"])
+        plan_data_subject_erasure(
+            self.session,
+            registry=registry,
+            row=request,
+            expected_revision=2,
+        )
+        executable_ids = [
+            action["action_id"]
+            for action in request.erasure_plan["actions"]
+            if action["executable"]
+        ]
+        execute_data_subject_erasure(
+            self.session,
+            registry=registry,
+            row=request,
+            expected_revision=3,
+            action_ids=executable_ids,
+        )
+        self.assertEqual("completed", request.status)
+
+        disabled = create_data_subject_request(
+            self.session,
+            tenant_id="tenant-1",
+            reference="DSAR-FILES-DISABLED",
+            request_kind="access",
+            subject=self.subject,
+            purpose="Verify disabled-module coverage.",
+            legal_basis="Article 15 GDPR",
+            due_at=None,
+            requested_by_account_id="privacy-officer",
+        )
+        search_data_subject_request(
+            self.session,
+            registry=_Registry(self.provider, files_active=False),
+            row=disabled,
+            expected_revision=1,
+        )
+
+        self.assertEqual(0, disabled.search_result["record_count"])
+        self.assertEqual(
+            [FILES_DSAR_CAPABILITY],
+            disabled.coverage["inactive_provider_capabilities"],
+        )
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_manifest_documentation.py

@@ -12,6 +12,7 @@ STATIC_TOPIC_IDS = {
     "files.workflow.share-managed-files",
     "files.workflow.delete-managed-files",
     "files.workflow.restore-retain-and-purge",
+    "files.privacy.data-subject-requests",
     "files.governed-connectors-and-provenance",
     "files.reference.integrity-recovery-and-fail-closed-transports",
     "files.reference.shared-storage-profile",
@@ -130,6 +131,14 @@ class FilesManifestDocumentationTests(unittest.TestCase):
         self.assertIn("preview hash", lifecycle.body)
         self.assertIn("recovery ledger", lifecycle.body)
 
+        privacy = self.topic("files.privacy.data-subject-requests")
+        self.assertEqual(("admin",), privacy.documentation_types)
+        self.assertIn("without raw file bytes", privacy.body)
+        self.assertIn("separately authorized Files purge", privacy.body)
+        self.assertTrue(
+            any("membership" in item for item in privacy.metadata["limitations"])
+        )
+
     def test_admin_topic_covers_policy_redaction_and_atomic_credential_deletion(
         self,
     ) -> None: