Flow-Research · Abiorh001 · Jul 8, 2026 · Jul 7, 2026 · Jul 7, 2026 · Jul 8, 2026
diff --git a/.agent-loop/LOOP_STATE.md b/.agent-loop/LOOP_STATE.md
@@ -4,18 +4,16 @@
 
 - Active initiative: `WS-POL-001` - Submission Artifact Policy Foundation
 - Active planning chunk: none
-- Active implementation chunk: `WS-POL-001-12`
-- Branch: `codex/ws-pol-001-12-project-setup-policy-visibility`
-- Status: `WS-POL-001-12` is implemented in PR #76 on branch
-  `codex/ws-pol-001-12-project-setup-policy-visibility`. Internal review is
-  complete, CodeRabbit feedback is being addressed in a follow-up commit, and
-  the chunk is waiting for human merge review after external-review fixes are
-  verified.
-- Last merged implementation SHA: `0729531`
-- Last merge commit: `5cec0e0`
-- Current gate: verify CodeRabbit fixes on PR #76, update trust evidence, then
-  wait for explicit human merge approval.
-- Next chunk: inactive until `WS-POL-001-12` receives human review and merge.
+- Active implementation chunk: `WS-POL-001-13`
+- Branch: `codex/ws-pol-001-13-task-context-apis`
+- Status: `WS-POL-001-13` is implementing worker-safe task work context,
+  exact submission requirements, and operator-only locked task provenance APIs
+  after PR #76 merged to `main`.
+- Last merged implementation SHA: `46e74de`
+- Last merge commit: `46e74de`
+- Current gate: implement Chunk 13, run deterministic proof, run required
+  internal reviewer tracks, then open a PR for human review.
+- Next chunk: inactive until `WS-POL-001-13` receives human review and merge.
 
 ## Operating Rule
 
@@ -115,3 +113,7 @@ blockchain, frontend, or agent-runtime behavior.
   actors before the next Terminal Benchmark live API drill.
 - `WS-POL-001-11` internal review evidence is tracked at `.agent-loop/initiatives/WS-POL-001-submission-artifact-policy-foundation/reviews/WS-POL-001-11-internal-review-evidence.md`.
 - `WS-POL-001-11` PR trust bundle is tracked at `.agent-loop/initiatives/WS-POL-001-submission-artifact-policy-foundation/reviews/WS-POL-001-11-pr-trust-bundle.md`.
+- PR #76 merged into `main` as `46e74de`; it implemented `WS-POL-001-12`
+  project setup-run and policy visibility APIs.
+- `WS-POL-001-13` started on branch `codex/ws-pol-001-13-task-context-apis`
+  after the user's explicit start signal.
diff --git a/.agent-loop/initiatives/WS-POL-001-submission-artifact-policy-foundation/STATUS.md b/.agent-loop/initiatives/WS-POL-001-submission-artifact-policy-foundation/STATUS.md
@@ -4,20 +4,22 @@
 
 `WS-POL-001-01`, `WS-POL-001-02`, `WS-POL-001-03`, `WS-POL-001-04`,
 `WS-POL-001-05`, `WS-POL-001-06`, `WS-POL-001-07`, `WS-POL-001-08`,
-`WS-POL-001-09`, `WS-POL-001-10`, and `WS-POL-001-11` are merged to `main`.
+`WS-POL-001-09`, `WS-POL-001-10`, `WS-POL-001-11`, and `WS-POL-001-12` are
+merged to `main`.
 The post-actor-registry Terminal Benchmark live API drill passed through real
-HTTP calls, but it required DB inspection for project setup outputs and locked
-context verification. `WS-POL-001-12` is implemented in PR #76 to expose the
-first project setup and project policy visibility APIs. `WS-POL-001-13` and
-`WS-POL-001-14` remain the next visibility/finalize chunks before rerunning the
-drill without DB inspection.
+HTTP calls, but it still needs task context visibility, submission finalize
+wording, and system-actor audit semantics before being accepted as a no-DB
+proof. `WS-POL-001-13` is active to expose worker-safe task context, exact
+submission requirements, and operator-only locked provenance. `WS-POL-001-14`
+remains the finalize/no-DB proof chunk before rerunning the Terminal Benchmark
+drill as accepted evidence.
 
 ## Active Chunk
 
-`WS-POL-001-12` is implemented on branch
-`codex/ws-pol-001-12-project-setup-policy-visibility` in PR #76. Internal
-review is complete, CodeRabbit feedback is being addressed, and the chunk is
-not merged until the user explicitly approves PR #76.
+`WS-POL-001-13` is active on branch
+`codex/ws-pol-001-13-task-context-apis`. It adds task work-context,
+submission-requirements, and operator locked-context APIs before the next
+Terminal Benchmark drill.
 
 ## Chunk Status
 
@@ -34,16 +36,15 @@ not merged until the user explicitly approves PR #76.
 | `WS-POL-001-09` | Merged | `codex/ws-pol-001-09-openai-agent-sdk-only` | 71 | Removes the production `local_fixture` project setup runtime and old runtime selector; keeps deterministic test behavior in explicit test-local fakes only. |
 | `WS-POL-001-10` | Merged | `codex/ws-pol-001-10-pre-submit-hardening` | 72 | Hardens duplicate guide-version conflicts, guide-create source snapshots, active-guide checker summaries, worker self-profile onboarding, and failed pre-submit audit evidence. |
 | `WS-POL-001-11` | Merged | `codex/ws-pol-001-11-actor-profile-registry-impl` | 74 | Implements local Workstream actor identity and actor profile registries for verified Flow actors before the next live API drill. |
-| `WS-POL-001-12` | PR open | `codex/ws-pol-001-12-project-setup-policy-visibility` | 76 | Adds project setup-run and project policy visibility APIs for setup runs, sufficiency reports, submission artifact policies, effective policy, and compiled project pre-submit checker policy; awaiting human merge review after external feedback fixes. |
-| `WS-POL-001-13` | Proposed | - | - | Add task work-context, worker submission-requirements, and operator locked-context APIs. |
+| `WS-POL-001-12` | Merged | `codex/ws-pol-001-12-project-setup-policy-visibility` | 76 | Adds project setup-run and project policy visibility APIs for setup runs, sufficiency reports, submission artifact policies, effective policy, and compiled project pre-submit checker policy. |
+| `WS-POL-001-13` | Active | `codex/ws-pol-001-13-task-context-apis` | - | Add task work-context, worker submission-requirements, and operator locked-context APIs. |
 | `WS-POL-001-14` | Proposed | - | - | Replace public submission lock with finalize, define system actor audit semantics, and rerun the Terminal Benchmark proof without DB inspection. |
 
 ## Blockers
 
 | Blocker | Owner | Next action |
 |---|---|---|
-| PR #76 human merge review | Human reviewer | Review and explicitly approve or request changes for `WS-POL-001-12`. |
-| Missing task context visibility APIs | Workstream | Implement `WS-POL-001-13`. |
+| Missing task context visibility APIs | Workstream | Implement and review `WS-POL-001-13`. |
 | Public lock wording and no-DB drill proof | Workstream | Implement `WS-POL-001-14` before rerunning Terminal Benchmark as accepted proof. |
 
 ## Follow-Ups

diff --git a/...on-artifact-policy-foundation/reviews/WS-POL-001-13-external-review-response.md b/...on-artifact-policy-foundation/reviews/WS-POL-001-13-external-review-response.md
@@ -0,0 +1,59 @@
+# External Review Response: WS-POL-001-13
+
+## PR
+
+PR #77: https://github.com/Flow-Research/workstream/pull/77
+
+## Chunk
+
+`WS-POL-001-13`
+
+## Source
+
+CodeRabbit
+
+## Summary
+
+CodeRabbit reported one valid test-maintainability nitpick and one PR
+description warning. No product, architecture, security, authorization, or API
+contract changes were requested.
+
+## External Findings
+
+| Source | Finding | Severity | Status | Response |
+|---|---|---:|---:|---|
+| CodeRabbit | The malformed locked-context regression test read `pre_submit_policy.compiled_bundle` after `session.delete()` and `flush()`. | Trivial | Addressed | Captured `original_compiled_bundle` before deleting the ORM row and used the saved bundle when constructing the malformed checker bundle. |
+| CodeRabbit | The PR description omitted explicit template sections for chunk, goal, changed behavior, acceptance proof, and test delta. | Warning | Addressed | Reshaped the PR trust bundle and PR body around the required review sections. |
+
+## Fix plan
+
+- Update `backend/tests/test_tasks.py` so the malformed locked-context test does
+  not dereference a deleted ORM instance.
+- Update the PR trust bundle with explicit review-template headings.
+- Record this external review response separately from internal review
+  evidence.
+- Update the GitHub PR body from the corrected trust bundle.
+
+## Out-of-scope items to defer
+
+None.
+
+## Evidence after fixes
+
+```bash
+cd backend && .venv/bin/pytest tests/test_tasks.py::test_submission_requirements_fail_closed_on_hash_consistent_malformed_policy_shape -q
+cd backend && .venv/bin/python -m ruff check tests/test_tasks.py
+python3 scripts/check_stale_workstream_wording.py
+python3 scripts/check_markdown_links.py
+git diff --check
+```
+
+Result summary:
+
+```text
+Focused pytest: 1 passed
+Ruff: All checks passed
+Stale wording: passed
+Markdown links: passed for changed Markdown files
+git diff --check: passed
+```
diff --git a/...on-artifact-policy-foundation/reviews/WS-POL-001-13-internal-review-evidence.md b/...on-artifact-policy-foundation/reviews/WS-POL-001-13-internal-review-evidence.md
@@ -0,0 +1,93 @@
+# Internal Review Evidence: WS-POL-001-13
+
+## Chunk
+
+WS-POL-001-13
+
+open sub-agent sessions: none
+
+valid findings addressed: yes
+
+## Reviewed Revision
+
+Reviewed code SHA: f533f1a572a38d4e8ecd34ff6316885c6c6b1016
+
+Reviewed at: 2026-07-08T02:09:55Z
+
+Reviewer run IDs: senior-engineering-final-019f3ecb-6123-7581-9b08-a1812b789d57, qa-test-final-019f3ecb-a018-7e02-81a7-3a7fd102a4ba, security-auth-final-019f3ecb-d5a9-7800-94b3-e4358974f0cd, product-ops-final-019f3ecc-0e2d-7e90-873a-af311a10fd59, architecture-final-019f3ecc-5034-7da0-a53e-1cb66880535b, docs-final-019f3ecc-87cf-75a3-b806-d694cf0f8114, reuse-dedup-final-019f3ed0-bdf1-7ae3-99ae-a33db36dea1c, test-delta-final-019f3ed0-f9d9-7590-a829-de92b9296105, senior-engineering-external-019f3f6d-441f-7340-9bcc-ef92155d956d, qa-test-external-019f3f6d-4986-7612-a7a7-9457a2182d0b, security-auth-external-019f3f6d-4fea-7890-a372-a261d2482b91, product-ops-external-019f3f6d-5758-79d3-a7d4-3d58933064da, docs-external-final-019f3f76-dd39-74f3-8cf3-d551908fa058, test-delta-external-019f3f6d-685a-7300-82ab-966c3c034976
+
+After the reviewed SHA, only evidence files changed.
+
+## Reviewed Change
+
+Scope:
+
+- Adds `GET /api/v1/tasks/{task_id}/work-context` for worker-safe task, project, guide, payment, review/revision version, and lifecycle context.
+- Adds `GET /api/v1/tasks/{task_id}/submission-requirements` for exact locked submission packet fields, artifacts, evidence, storage rules, packaging, hash algorithm, and attestation terms.
+- Adds `GET /api/v1/tasks/{task_id}/locked-context` for `admin` and `project_manager` locked provenance.
+- Validates task-stamped guide source snapshot, effective project policy, project pre-submit checker, post-submit checker, review, revision, and payment context before response projection.
+- Keeps worker-facing responses free of private source/import refs, assignment provenance, raw compiled bundles, checker configs, setup internals, source snapshot hashes, and policy provenance hashes.
+- Redacts existing worker-visible `GET /tasks/{task_id}` responses so workers cannot bypass `work-context` redaction.
+- Fails closed with structured `task_locked_context_invalid` when locked context is missing, stale, hash-mismatched, or hash-consistent but malformed.
+- Extends the real API contract drill to call the three task-context APIs through HTTP.
+
+## Reviewer Results
+
+| Reviewer | Result | Blocking findings | Notes |
+|---|---:|---|---|
+| senior engineering | PASS | None | Confirmed router/service boundaries, route-specific domain schemas, fail-closed malformed policy behavior, and worker redaction. |
+| QA/test | PASS | None | Confirmed acceptance criteria, exact requirements including `package_hash`, v1 lock after v2 activation, and live API drill coverage. |
+| security/auth | PASS WITH LOW RISKS | None | Confirmed no private values/hashes/configs leak. Low residual risk: worker-facing structured errors include internal field names without values. |
+| product/ops | PASS | None | Confirmed worker and operator workflows, existing task read redaction, and no payment/reputation behavior expansion. |
+| architecture | PASS | None | Confirmed stamped-context reads, no current-guide recompute, operator-only locked context, and allowed-file scope. |
+| docs | PASS | None | Confirmed glossary, architecture, operations, and roadmap docs align with the new API surfaces. |
+| reuse/dedup | PASS WITH LOW RISKS | None | Low residual risk: required packet field constants mirror `SubmissionCreate`; future shared extraction can avoid drift. |
+| test delta | PASS WITH LOW RISKS | None | Confirmed strengthened tests and no skips/removals. Low residual risk: locked-context response shape is sampled rather than fully enumerated. |
+
+## Valid Findings Addressed
+
+- Replaced broad worker `TaskResponse` usage in `work-context` with `TaskWorkerTaskContext`.
+- Removed private source/import/creator/assignment fields from worker `work-context` task data.
+- Stopped exposing mutable same-version review/revision/payment row details in worker work context; response now uses stamped guide/payment values and guide-version references.
+- Added route-specific domain error schemas so submission creation documents only `pre_submission_checker_failed` and task-context endpoints document only `task_locked_context_invalid`.
+- Added `package_hash` to the exact worker-facing required packet field list.
+- Removed `forbidden_artifacts.source` from worker-facing submission requirements.
+- Redacted private source/import/creator/assignment fields from existing worker-visible `GET /tasks/{task_id}` responses.
+- Added explicit locked effective-policy projection validation so malformed hash-consistent policy shapes return structured `task_locked_context_invalid`.
+- Centralized locked-context required fields and reused the shared missing-field helper.
+- Added tests for worker-safe redaction, operator-only locked context, exact submission requirements, malformed policy fail-closed behavior, same-version policy mutation, and v1 locked context after v2 guide activation.
+- Updated the real API contract drill to call `work-context`, `submission-requirements`, and `locked-context`.
+- Addressed CodeRabbit's test-maintainability nitpick by capturing the compiled
+  bundle before deleting the ORM row in the malformed locked-context
+  regression.
+
+## Commands Run
+
+```bash
+cd backend && .venv/bin/python -m ruff check app tests scripts
+python3 scripts/check_stale_workstream_wording.py
+python3 scripts/check_markdown_links.py
+git diff --check
+rg -n "task binding|task-owned policy|direct DB|DB inspection" docs backend examples .agent-loop
+cd backend && .venv/bin/pytest tests/test_tasks.py::test_worker_task_response_redacts_locked_policy_hashes tests/test_tasks.py::test_assigned_worker_submits_v1_and_task_moves_to_submitted tests/test_tasks.py::test_task_context_apis_return_worker_requirements_and_operator_provenance tests/test_tasks.py::test_submission_requirements_fail_closed_on_hash_consistent_malformed_policy_shape -q
+cd backend && WORKSTREAM_DATABASE_URL=postgresql+asyncpg://workstream:workstream@localhost:5433/workstream_test .venv/bin/python scripts/api_contract_e2e.py
+cd backend && .venv/bin/pytest tests/test_tasks.py
+```
+
+Results:
+
+- Ruff: passed.
+- Stale wording scan: passed.
+- Markdown link check: passed for 7 changed Markdown files.
+- Diff whitespace check: passed.
+- Stale model/no-DB scan: expected references only in chunk contracts, status/decision docs, and the documented future no-DB proof path; no active per-task policy wording was found.
+- Focused task context and worker redaction regressions: 4 passed in 43.18s.
+- API contract real API E2E: passed and called `work-context`, `submission-requirements`, and `locked-context` through HTTP.
+- Full task suite: 81 passed in 743.67s.
+
+## Remaining Risks
+
+- Worker-facing locked-context errors include internal field names such as locked context field identifiers. They do not expose values, hashes, private refs, compiled bundles, checker configs, or source material. A future UX/security hardening pass can map worker errors to public categories while preserving exact operator diagnostics.
+- `SUBMISSION_CREATE_REQUIRED_PACKET_FIELDS` mirrors `SubmissionCreate` required fields. It is centralized for this projection, but a future cleanup can derive it from schema metadata or move the contract beside the schema to reduce drift.
+- The task context loader repeats some locked pre-submit validation concepts also used by checker execution. The current direction is acceptable because this API needs task-specific domain errors and richer context; if another reader appears, extract a shared locked-context validator.
+- The next Terminal Benchmark no-DB drill is still assigned to `WS-POL-001-14`.