docs(rev): split guide chronology and task locking contracts#156
Conversation
📝 WalkthroughWalkthroughWS-REV-001 parent 02A is now a planning-only split. Executable work is divided into 02A1, 02A3, and 02A4, while 02A2 remains deferred. Lifecycle documentation, dependency baselines, review evidence, downstream prerequisites, and explicit-start metadata were updated. ChangesREV-02A lifecycle planning
Estimated code review effort: 3 (Moderate) | ~25 minutes Possibly related PRs
Suggested reviewers: 🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In
@.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/SOURCE_MANIFEST.md:
- Around line 225-238: Update the 02A split review-base records to distinguish
the historical AUTH-09D-A/0026 review base from the actual trusted-main start at
migration 0027_contributor_foundation. Use the full 0027 commit SHA consistently
in the later post-rebase review-base section, ensuring child migration
allocation starts from that trusted head rather than the stale 0026 commit.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro Plus
Run ID: b53714e9-bfc4-4d28-b391-4fc296fdfe0d
📒 Files selected for processing (21)
.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/CHUNK_MAP.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/CONFORMANCE_MATRIX.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/DECISIONS.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/DISCOVERY.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/PLAN.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/REVIEW_LOG.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/RISKS.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/SOURCE_MANIFEST.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/STATUS.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/TEST_DESIGN_WS-REV-001-02.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02-review-policy-task-alignment.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02A-guide-activation-sequence.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02A1-project-setup-publication-fence.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02A2-prepared-guide-reactivation.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02A3-guide-activation-chronology.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02A4-task-guide-triplet-screening.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02B-review-policy-task-lifecycle.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02C-submission-attribution-lineage.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/reviews/WS-REV-001-02A-internal-review-evidence.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/reviews/WS-REV-001-02A-pr-trust-bundle.md.agent-loop/merge-intents/WS-REV-001-02A.json
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In
@.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/reviews/WS-REV-001-02A-external-review-response.md:
- Around line 48-49: The external-review response document must identify the
stale-SHA result as historical or replace it with final evidence for the
repaired candidate. In
.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/reviews/WS-REV-001-02A-external-review-response.md
lines 48-49, update the evidence status accordingly; in
.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/reviews/WS-REV-001-02A-pr-trust-bundle.md
lines 102-116, explicitly block merge until current-head GitHub/CodeRabbit
rechecks and evidence rebinding are complete, or record their completed results.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro Plus
Run ID: 9cb87c74-33c7-4eb3-a55b-f02f2fc7bf96
📒 Files selected for processing (5)
.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/REVIEW_LOG.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/SOURCE_MANIFEST.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/reviews/WS-REV-001-02A-external-review-response.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/reviews/WS-REV-001-02A-internal-review-evidence.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/reviews/WS-REV-001-02A-pr-trust-bundle.md
🚧 Files skipped from review as they are similar to previous changes (3)
- .agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/SOURCE_MANIFEST.md
- .agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/REVIEW_LOG.md
- .agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/reviews/WS-REV-001-02A-internal-review-evidence.md
1c2706a to
b4a2b07
Compare
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In
@.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/CONFORMANCE_MATRIX.md:
- Line 14: Update the “Revision context” entry in CONFORMANCE_MATRIX.md to
include deferred chunk 02A2, preserving the existing ownership and invariant
descriptions. Ensure prepared If-Match-protected superseded-guide reactivation
is explicitly owned by 02A2 or clearly reassigned to another chunk.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro Plus
Run ID: d1881882-4ae0-460d-b9ba-1b4bd1124dec
📒 Files selected for processing (22)
.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/CHUNK_MAP.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/CONFORMANCE_MATRIX.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/DECISIONS.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/DISCOVERY.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/PLAN.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/REVIEW_LOG.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/RISKS.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/SOURCE_MANIFEST.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/STATUS.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/TEST_DESIGN_WS-REV-001-02.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02-review-policy-task-alignment.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02A-guide-activation-sequence.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02A1-project-setup-publication-fence.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02A2-prepared-guide-reactivation.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02A3-guide-activation-chronology.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02A4-task-guide-triplet-screening.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02B-review-policy-task-lifecycle.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02C-submission-attribution-lineage.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/reviews/WS-REV-001-02A-external-review-response.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/reviews/WS-REV-001-02A-internal-review-evidence.md.agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/reviews/WS-REV-001-02A-pr-trust-bundle.md.agent-loop/merge-intents/WS-REV-001-02A.json
🚧 Files skipped from review as they are similar to previous changes (16)
- .agent-loop/merge-intents/WS-REV-001-02A.json
- .agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02B-review-policy-task-lifecycle.md
- .agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02C-submission-attribution-lineage.md
- .agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/RISKS.md
- .agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02A3-guide-activation-chronology.md
- .agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/CHUNK_MAP.md
- .agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02-review-policy-task-alignment.md
- .agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02A4-task-guide-triplet-screening.md
- .agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/reviews/WS-REV-001-02A-internal-review-evidence.md
- .agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/DECISIONS.md
- .agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/SOURCE_MANIFEST.md
- .agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/PLAN.md
- .agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02A1-project-setup-publication-fence.md
- .agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/TEST_DESIGN_WS-REV-001-02.md
- .agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/REVIEW_LOG.md
- .agent-loop/initiatives/WS-REV-001-review-revision-lifecycle/chunks/WS-REV-001-02A-guide-activation-sequence.md
PR Trust Bundle: WS-REV-001-02A
Chunk
WS-REV-001-02A- Guide Chronology And Task Locking Split.Goal
Replace one oversized L1 runtime contract with three independently reviewable
and explicitly started children, without implementing runtime behavior.
Human-approved intent
Preserve one Project Guide pipeline, exact Task-bound guide context,
canonical-human approval provenance, immutable chronology, and the existing
separation between checker remediation and Review-rooted human revision. Do not
start any runtime child from this planning parent.
What changed
guide activation chronology, and 02A4 Task guide triplet screening.
source manifest, test design, status, review log, and downstream gates.
Why it changed
Required L1 preimplementation review found that the former contract crossed
three independent PostgreSQL concurrency and migration boundaries: the complete
Project/setup writer fence, canonical-human guide chronology, and Task guide
triplet screening. Those boundaries could not receive credible migration,
direct-SQL, race, rollback, downgrade, coverage, and human review proof in one
PR.
Design chosen
02A1 serializes every current setup writer through a Project-first publication
fence. 02A3 builds immutable per-Project activation chronology and canonical
human provenance on that fence. 02A4 then stamps and protects the exact Task
guide triplet after chronology exists. The later 02A2 remains separately gated
after chunk 08 for hidden prepared superseded-guide reactivation.
Alternatives rejected
Scope control
The full PR changes 21 Markdown files under the REV initiative and one
merge-intent JSON file. This includes the internal evidence, trust bundle, and
external-review response added after the original 19-path planning candidate.
It changes no backend code, model, schema, migration, service, route, worker,
test, dependency, CI workflow, frozen reference specification, active product
document, or AUTH/ART/CON owner artifact.
Product behavior
This PR activates no product behavior. Current guide activation continues to
reject an already-active candidate. Task guide storage and screening remain
unchanged. Stored Review decisions remain only
accept,needs_revision, andreject, and adjudication remains deferred beyond v0.1.Acceptance criteria proof
The parent records exact merged AUTH dependencies and the sole migration head.
Its historical start head was
0027_contributor_foundation; the conflict-freepublication rebase onto ART PR #154 advances the current sole head to
0028_artifact_admissionwithout adding a Project/setup writer.Every child specifies allowed and prohibited files, lock order, race and
rollback proof, coverage floors, required reviewers, human focus, and a stop
condition. Migration-owning children 02A3 and 02A4 additionally specify their
database objects, migration refusal, direct-SQL parity, and downgrade rules.
The merge intent names only same-initiative successor 02A1 and requires a
separate human start.
Tests/checks run
All four stale-contract scanners, Markdown links, loop-memory state, 88 agent
gate tests, docstring coverage at 90.5 percent, one-head Alembic verification,
schema-v2 merge-intent validation, scope checks, ancestry checks, and diff
integrity passed on the planning candidate.
Test delta
No executable test changed. No assertion, skip, xfail, test command, coverage
floor, or CI gate was removed or weakened.
CI integrity
No CI or package-script file changed. The repository-wide 78 percent baseline
and independent 90 percent minimum for future new or materially changed backend
subsystems remain explicit in the executable child contracts.
Reviewer results
Senior engineering, QA/test, security/auth, product/ops, architecture, docs,
reuse/dedup, test delta, and CI integrity all pass on rebased planning SHA
73a6c31ffb54a11cd22992dd3b6d0de5413d7e05against ART-merged trusted base44f2467cedc266d2efe261119cfff436ac6b7715. The circuit breaker approved theplanning-only size exception because every executable boundary is split and
separately gated.
External review
After the prior review cycle passed, ART PR #154 merged and PR #156 rebased
without conflict onto trusted main
44f2467c. This rewrote every branch SHA;all prior GitHub/CodeRabbit results are historical. The rebased candidate has
now passed exact-SHA internal review and evidence rebinding; it must still be
force-pushed and pass fresh final-head external checks.
On prior pushed head
8f12f656, GitHub Backend and Agent Gates passed, andCodeRabbit passed with one actionable source-manifest finding: a historical
PLAN2 AUTH-09D-A/
0026review-base sentence could be mistaken for the laterparent 02A start. The repair labels that snapshot as historical and restates the
full 02A trusted-main SHA,
0027_contributor_foundationhead, andthen-current-head child allocation rule.
Fresh deterministic and internal review evidence passes for the repaired
candidate. At evidence head
ebb8db88, Agent Gates passed, Backend remainedpending, and CodeRabbit's evidence-status comment required this repair. Merge is
blocked until all GitHub and CodeRabbit checks pass on the final pushed head with
no unresolved actionable comment. Human review remains pending.
On the ART-rebased head, CodeRabbit also found that the Revision-context
conformance row omitted deferred owner 02A2. The row now includes 02A2 and its
prepared
If-Match-protected superseded-guide reactivation proof; runtime andthe explicit-start gate are unchanged. Exact-SHA internal review and evidence
rebinding pass for that repaired candidate.
Remaining risks
Historical chronology or Task context ambiguity must fail closed in future
migrations. ART's generic admission foundation is merged, while remaining ART
review capabilities and CON capabilities remain dependency-owned future work. All
24 planned REV action dependencies remain unavailable. Later duration and human
revision exhaustion semantics still require explicit human decisions.
Follow-up work
After merge, automated loop memory may name 02A1 with an explicit-start gate.
02A1 must refresh trusted main, the complete writer inventory, migration head,
plan review, and reviewer routing before any implementation. 02A3, 02A4, 02A2,
and 02B remain separately gated successors.
Human review focus
Review the split boundaries and order, complete writer-fence inventory,
canonical-human ownership, immutable chronology, migration refusal/downgrade
safety, Task triplet integrity, and the absence of runtime changes.
Human merge ownership
Only the user may approve and merge this specific PR. Merge does not authorize
02A1 or any other runtime implementation. Do not merge while any current-head
GitHub or CodeRabbit check is pending or failed, or while an actionable review
comment remains unresolved.