project-loop-harness 0.5.3__tar.gz → 0.5.4__tar.gz

{project_loop_harness-0.5.3/src/project_loop_harness.egg-info → project_loop_harness-0.5.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: project-loop-harness
-Version: 0.5.3
+Version: 0.5.4
 Summary: Local control plane for agentic project development loops
 Author: Project Loop Harness Contributors
 License: MIT

project_loop_harness-0.5.4/agent-tasks/0203-v053-release-publication.md

@@ -0,0 +1,56 @@
+# 0203: v0.5.3 release publication
+
+- **Status:** Done; published and publicly verified
+- **Milestone:** v0.5.3 Evidence Integrity and Harness Feedback
+- **Priority:** P0
+- **Size:** S
+- **Dependency:** completed 0202 local release candidate; explicit owner publication instruction
+- **Project Loop:** Goal `G-0061`, Task `T-0125`, Feature `F-0067`, Story `US-0065`, Test `TC-0138`
+- **DB schema:** remains 8
+
+## Goal
+
+Publish the reviewed v0.5.3 release commit and verify the complete public
+artifact chain without converting artifact publication into an adoption claim.
+
+## Scope
+
+1. Push the reviewed release commit and require its GitHub CI matrix to pass.
+2. Create and push an annotated `v0.5.3` tag at that exact commit.
+3. Publish the GitHub Release and verify release-triggered Trusted Publishing.
+4. Verify PyPI metadata, exact artifact hashes, and a clean public install.
+5. Replace the editable pipx source with the released PyPI version and record
+   the immutable closeout evidence.
+
+## Invariants
+
+- No schema migration, dependency, telemetry, hosted state, provider execution,
+  launch post, or unrelated repair is added.
+- Publication is engineering evidence, not evidence of external adoption.
+- Existing `.claude` session state and unrelated worktree files are excluded
+  from release commits.
+
+## Acceptance
+
+1. `main`, annotated tag, GitHub Release target, package version, and public
+   PyPI artifacts all resolve to v0.5.3.
+2. Release-commit CI and Trusted Publishing succeed.
+3. Public wheel and sdist bytes match PyPI-reported SHA-256 digests.
+4. A clean PyPI install passes version, contract, init, strict validation,
+   audit, and render checks.
+5. Pipx runs the non-editable public v0.5.3 package.
+
+## Completion evidence
+
+- Release commit: `6e49ac43986e8b965f1777eef9a88f7c73236ef6`
+- Annotated tag object: `b5eb54666e8b537b17fd459bd1d9aa5cba94db89`
+- GitHub Release: `v0.5.3`, published `2026-07-20T06:30:42Z`
+- Green release-commit CI run: `29721069992`
+- Trusted Publishing run: `29721897474`, success
+- Public wheel SHA-256:
+  `0ce97a13c6deedf6525a8487aaba8080744e303c96845ee22028ad5bd623c54f`
+- Public sdist SHA-256:
+  `448166e9ed78c68ccba2c13dd4d9a20038c2e5e105b717ca97d5ebb02b07eb0a`
+- Clean public install, Gap Report contract, strict validation, audit, render,
+  and pipx replacement: passed
+- `docs/evidence/0203-v053-publication-closeout.md`

project_loop_harness-0.5.4/agent-tasks/0204-cli-stage3-entity-handlers.md

@@ -0,0 +1,28 @@
+# 0204: CLI Stage 3 entity lifecycle handlers
+
+- **Status:** Done; implemented and verified
+- **Milestone:** Post-v0.5.3 maintainability
+- **Priority:** P1
+- **Size:** M
+- **Dependency:** 0188 and the frozen split contract
+- **DB schema:** remains 8
+
+## Goal
+
+Extract Goal, Task, Feature, Story, Test, and Defect CLI orchestration from
+`cli.py` without changing observable behavior or mutation ownership.
+
+## Acceptance
+
+1. Parser definitions and service-layer transactions remain in their current
+   owners.
+2. Exact JSON/text output, errors, exit codes, events, and zero-trace
+   rejections are characterized and preserved.
+3. Targeted tests, Skill examples, distribution tests, Ruff, and full pytest
+   pass.
+
+## Completion evidence
+
+- `docs/evidence/0204-cli-stage3-entity-handlers.md`
+- Implementation commit: `d7d38ea`
+- Full regression: 1,164 passed, 1 skipped

project_loop_harness-0.5.4/agent-tasks/0205-cli-stage3-execution-evidence-handlers.md

@@ -0,0 +1,27 @@
+# 0205: CLI Stage 3 execution and Evidence handlers
+
+- **Status:** Done; implemented and verified
+- **Milestone:** Post-v0.5.3 maintainability
+- **Priority:** P1
+- **Size:** M
+- **Dependency:** 0204
+- **DB schema:** remains 8
+
+## Goal
+
+Extract Loop, Workflow, Jobs, Agent, Evidence, Verification, Decision,
+Escalation, and checkpoint CLI orchestration while preserving every mutation,
+event, outbox, artifact, and rejection contract.
+
+## Acceptance
+
+1. Services retain transaction and event ownership.
+2. Handler parity and failure zero-trace tests cover each moved family.
+3. Targeted tests, Skill examples, distribution tests, Ruff, and full pytest
+   pass.
+
+## Completion evidence
+
+- `docs/evidence/0205-cli-stage3-execution-evidence-handlers.md`
+- Implementation commit: `dd9a327`
+- Full regression: 1,169 passed, 1 skipped

project_loop_harness-0.5.4/agent-tasks/0206-cli-stage3-control-profile-handlers.md

@@ -0,0 +1,22 @@
+# 0206: CLI Stage 3 control and Profile handlers
+
+- **Status:** Complete
+- **Milestone:** Post-v0.5.3 maintainability
+- **Priority:** P1
+- **Size:** M
+- **Dependency:** 0205
+- **DB schema:** remains 8
+
+## Goal
+
+Extract Profile, contract, Evidence Set, completion, Work Brief, Gap Report,
+route/policy, initialization, audit/repair, context, index/eval, export/report,
+and remaining CLI orchestration from `cli.py`.
+
+## Acceptance
+
+1. Existing human gates, dry-run boundaries, artifacts, and update-check
+   behavior remain unchanged.
+2. Characterization tests pin moved JSON/text/error branches.
+3. Targeted tests, Skill examples, distribution tests, Ruff, and full pytest
+   pass.

project_loop_harness-0.5.4/agent-tasks/0207-command-service-responsibility-split.md

@@ -0,0 +1,21 @@
+# 0207: Command service responsibility split
+
+- **Status:** Complete
+- **Milestone:** Post-v0.5.3 maintainability
+- **Priority:** P1
+- **Size:** M
+- **Dependency:** 0206
+- **DB schema:** remains 8
+
+## Goal
+
+Turn `pcl.commands` into a compatibility facade over responsibility-specific
+domain, next-action routing, and finish-planning modules.
+
+## Acceptance
+
+1. Existing imports such as `pcl.commands.create_goal`, `loop_status`,
+   `build_next_action`, and `to_pretty_json` remain valid.
+2. Query ordering, transactions, event payloads, routing priority, and finish
+   plans remain byte-for-byte compatible where serialized.
+3. Direct service tests, next-action tests, Ruff, and full pytest pass.

project_loop_harness-0.5.4/agent-tasks/0208-cli-stage4-parser-construction.md

@@ -0,0 +1,20 @@
+# 0208: CLI Stage 4 parser construction split
+
+- **Status:** Complete
+- **Milestone:** Post-v0.5.3 maintainability
+- **Priority:** P1
+- **Size:** M
+- **Dependency:** 0207
+- **DB schema:** remains 8
+
+## Goal
+
+Split parser construction by command family while retaining one public
+`build_parser()` facade and the existing source/wheel/sdist entry points.
+
+## Acceptance
+
+1. Commands, flags, defaults, choices, help text, aliases, and parser errors
+   remain unchanged.
+2. Skill example parsing and complete help smoke checks pass.
+3. Distribution tests, Ruff, full pytest, and source/wheel/sdist smoke pass.

project_loop_harness-0.5.4/agent-tasks/0209-refactoring-integrated-verification.md

@@ -0,0 +1,23 @@
+# 0209: Refactoring integrated verification
+
+- **Status:** Complete
+- **Milestone:** Post-v0.5.3 maintainability
+- **Priority:** P1
+- **Size:** S
+- **Dependency:** 0208
+- **DB schema:** remains 8
+
+## Goal
+
+Verify the complete staged refactor as one integrated source and distribution
+surface, record durable Evidence, and close the refactoring Goal.
+
+## Acceptance
+
+1. Ruff, targeted tests, Skill examples, distribution tests, and full pytest
+   pass on the final revision.
+2. Doctor, strict validation, audit check, render, and diff check complete with
+   no new findings.
+3. Source, wheel, and sdist command/help smoke tests pass.
+4. PCL tasks are terminal, a completion packet is emitted, and the Goal is
+   closed with that packet Evidence.

project_loop_harness-0.5.4/agent-tasks/0210-v054-local-release-preparation.md

@@ -0,0 +1,52 @@
+# 0210: v0.5.4 local release preparation
+
+- **Status:** Complete
+- **Milestone:** v0.5.4 Maintainability
+- **Priority:** P0
+- **Size:** M
+- **Dependency:** 0209 Refactoring integrated verification
+- **Project Loop:** Goal `G-0063`, Task `T-0133`, Feature `F-0068`, Story `US-0066`, Test `TC-0139`
+- **DB schema:** remains 8
+
+## Goal
+
+Prepare a reviewable local v0.5.4 release candidate containing the completed
+post-v0.5.3 behavior-preserving CLI and command-service refactor. Keep remote
+CI and publication as separate, explicitly authorized operations.
+
+## Scope
+
+1. Align package, runtime, MCP fixture, baseline fixture, task index, and
+   release-note surfaces on v0.5.4.
+2. Describe only the behavior-preserving refactor included since public
+   v0.5.3; do not claim new command or runtime behavior.
+3. Run source QA, MCP conformance, advisory retrieval evaluation,
+   scratch-project validation/render, wheel and sdist contracts, metadata
+   checks, and a clean wheel-install smoke.
+4. Record artifact hashes, exact verification results, and known repository
+   audit findings as durable Evidence.
+5. Commit the complete local release candidate without pushing, tagging,
+   publishing, announcing, or changing pipx.
+
+## Invariants
+
+- No Git push or tag, GitHub Release, PyPI/TestPyPI upload, pipx mutation, or
+  external announcement.
+- No intentional command, flag, output, exit-code, transaction, event,
+  Evidence, human-gate, schema, dependency, or generated-artifact change.
+- Existing `.claude`, `.playwright-cli`, `.work`, and Project Loop lock files
+  remain outside the release commit.
+- Public v0.5.3 records remain historical v0.5.3 records; they are not renamed.
+
+## Acceptance
+
+1. `pyproject.toml`, `pcl.__version__`, CLI/MCP runtime output, wheel, and
+   sdist metadata agree on `0.5.4`.
+2. Ruff, the full pytest suite, optional MCP conformance, and the advisory
+   retrieval evaluation pass from the canonical source checkout.
+3. A fresh scratch project passes init, strict validation, clean audit, and
+   render using the source checkout.
+4. Wheel and sdist build, `twine check`, extracted-sdist contracts, and a
+   clean-wheel runtime/init/strict-validation/audit/render smoke pass.
+5. SHA-256 hashes, file sizes, candidate boundary, known warnings, and the
+   publication boundary are reviewable in a new write-once Evidence note.

{project_loop_harness-0.5.3 → project_loop_harness-0.5.4}/agent-tasks/README.md

@@ -120,6 +120,14 @@ implement from the spec file, not from prompt summaries.
 | 0200 | `gap-report/v1` Harness Gap Evidence | Harness Engineering Feedback Loop | P0 | done |
 | 0201 | `gap-report/v1` Integrity Hardening | Harness Engineering Feedback Loop | P1 | done |
 | 0202 | v0.5.3 local release preparation | v0.5.3 Evidence Integrity and Harness Feedback | P0 | done; local RC ready, not published |
+| 0203 | v0.5.3 release publication | v0.5.3 Evidence Integrity and Harness Feedback | P0 | done; published and publicly verified |
+| 0204 | CLI Stage 3 entity handler extraction | post-v0.5.3 maintainability | P1 | done |
+| 0205 | CLI Stage 3 execution/Evidence handler extraction | post-v0.5.3 maintainability | P1 | done |
+| 0206 | CLI Stage 3 control/Profile handler extraction | post-v0.5.3 maintainability | P1 | done |
+| 0207 | Command service responsibility split | post-v0.5.3 maintainability | P1 | done |
+| 0208 | CLI Stage 4 parser construction split | post-v0.5.3 maintainability | P1 | done |
+| 0209 | Refactoring integrated verification | post-v0.5.3 maintainability | P1 | done |
+| 0210 | v0.5.4 local release preparation | v0.5.4 Maintainability | P0 | done; local RC ready, not published |
 
 v0.3.0 dispatch order: **0113 + 0114 in parallel** (independent; different
 `evidence.py` surfaces) → **0108** (needs 0113 merged) → **0115** (freezes the
@@ -200,8 +208,20 @@ external first-use cohort remains open and no adoption claim is made. Task
 owner's separate publication instruction.
 
 v0.5.3 packages event-anchored strict Evidence resolution and the bounded
-`gap-report/v1` feedback layer. Task **0202** prepares and verifies the local
-release candidate; tagging, pushing, and public publication remain separate.
+`gap-report/v1` feedback layer. Task **0202** prepared and verified the local
+release candidate; task **0203** published and independently verified the
+GitHub Release, Trusted Publishing run, PyPI artifacts, and clean public install.
+
+Post-v0.5.3 maintainability completes the frozen behavior-preserving split:
+**0204** entity lifecycle handlers -> **0205** execution/Evidence handlers ->
+**0206** control/Profile handlers -> **0207** `commands.py` responsibility
+split -> **0208** parser construction -> **0209** integrated verification.
+See `docs/plan-refactoring-completion.md`.
+
+v0.5.4 packages that completed behavior-preserving split as a patch release.
+Task **0210** aligns version and distribution surfaces, records local artifact
+Evidence, and keeps push, tag, GitHub Release, PyPI, and pipx changes behind a
+separate publication authorization.
 
 ## Planned next (see growth plan for scope)
 
@@ -218,7 +238,9 @@ release candidate; tagging, pushing, and public publication remain separate.
 | v0.5.1 | Trace & Efficient Handoff: source-bound claims-not-facts handoff plus controlled cross-session/runtime resume evidence |
 | post-v0.5.1 | Behavior-preserving staged CLI split; 0184 presentation, then 0185 guide, 0186 loop status, 0187 doctor, and 0188 report-artifact handlers |
 | v0.5.2 | 0189 config-ready adoption and frozen cohort → 0190 participant kit and deterministic proof evaluator → five-user outcome cohort → observed-friction repair |
-| v0.5.3 | strict copied-Evidence resolution → 0200 Gap Report contract → 0201 integrity hardening → 0202 local release candidate |
+| v0.5.3 | published and publicly verified; strict copied-Evidence resolution → 0200 Gap Report contract → 0201 integrity hardening → 0202 local RC → 0203 publication closeout |
+| post-v0.5.3 | complete frozen CLI/commands refactor: 0204 → 0205 → 0206 → 0207 → 0208 → 0209 |
+| v0.5.4 | behavior-preserving CLI/commands maintainability release → 0210 local RC → separately authorized publication |
 
 Everything with an ID below 0102 is completed design history; see `TASKS.md`
 for the one-line summary of each.

project_loop_harness-0.5.4/docs/evidence/0203-v053-publication-closeout.md

@@ -0,0 +1,74 @@
+# 0203 v0.5.3 publication closeout
+
+**Verified:** 2026-07-20
+
+**Outcome:** published and publicly verified
+
+## Immutable source chain
+
+- Release commit: `6e49ac43986e8b965f1777eef9a88f7c73236ef6`
+- Annotated tag object: `b5eb54666e8b537b17fd459bd1d9aa5cba94db89`
+- Tag target: `6e49ac43986e8b965f1777eef9a88f7c73236ef6`
+- GitHub Release:
+  `https://github.com/mocchalera/project-loop-harness/releases/tag/v0.5.3`
+- Release published at: `2026-07-20T06:30:42Z`
+- Release is neither draft nor prerelease.
+
+## CI and publication
+
+- Release-commit CI run `29721069992`: success across Python 3.10, 3.11,
+  3.12, and 3.13; Ubuntu/Windows MCP conformance; Windows CLI smoke; build;
+  installed CLI smoke; and sdist contracts.
+- Release-triggered Trusted Publishing run `29721897474`: success.
+- Build-distributions job `88286464842`: success.
+- Publish-to-PyPI job `88286529538`: success.
+- TestPyPI was intentionally skipped by the production Release workflow.
+
+GitHub Actions emitted its platform warning that `actions/checkout@v4`,
+`actions/setup-python@v5`, and `actions/upload-artifact@v4` target deprecated
+Node.js 20 and were forced onto Node.js 24. The jobs passed; upgrading these
+actions remains follow-up maintenance rather than a v0.5.3 publication blocker.
+
+## Public PyPI artifacts
+
+PyPI reports `project-loop-harness 0.5.3`, `Requires-Python >=3.10`, one wheel,
+and one sdist. Independently downloaded bytes matched the PyPI JSON digests.
+
+| Artifact | Size | Uploaded | SHA-256 |
+| --- | ---: | --- | --- |
+| `project_loop_harness-0.5.3-py3-none-any.whl` | 504394 | `2026-07-20T06:31:34.657058Z` | `0ce97a13c6deedf6525a8487aaba8080744e303c96845ee22028ad5bd623c54f` |
+| `project_loop_harness-0.5.3.tar.gz` | 1462126 | `2026-07-20T06:31:36.400376Z` | `448166e9ed78c68ccba2c13dd4d9a20038c2e5e105b717ca97d5ebb02b07eb0a` |
+
+## Clean public install
+
+A new Python 3.13 venv at `/tmp/pcl-v053-public-verify.DPhCjr` installed
+`project-loop-harness==0.5.3` from PyPI with `--no-cache-dir` and no
+`PYTHONPATH`:
+
+- CLI, imported package, and installed metadata all reported `0.5.3`;
+- the packaged `gap-report/v1` validator accepted the minimal fixture;
+- non-empty-project init dry-run and apply passed;
+- strict doctor and strict validation returned zero findings;
+- audit was clean with 9 matching SQLite/JSONL events and zero anomalies;
+- dashboard render passed.
+
+The pipx installation was replaced with the explicit public spec
+`project-loop-harness==0.5.3`; both `pcl` and `pcl-mcp` are registered and the
+installed package is no longer editable from the local repository. Pipx still
+reports an invalid interpreter for unrelated package `haconiwa`; no repair was
+attempted because it is outside this release.
+
+## Repository harness state
+
+- `pcl validate --strict --json`: `ok: true`; 3 active and 26 historical
+  warnings from pre-existing Evidence/lifecycle history.
+- `pcl audit check --json`: 55 pre-existing human-review anomalies, with no
+  repairable or unsupported anomalies. This publication does not rewrite
+  historical Evidence or repair unrelated records.
+- Validation ran before the final dashboard render.
+
+## Claim boundary
+
+This release proves artifact integrity and engineering verification only. It
+does not establish external adoption, activation thresholds, or seven-day
+reuse. The prepared external first-use cohort remains separate evidence work.

project_loop_harness-0.5.4/docs/evidence/0204-cli-stage3-entity-handlers.md

@@ -0,0 +1,34 @@
+# 0204 CLI Stage 3 entity handlers evidence
+
+## Result
+
+Goal, Task, Feature, Story, Test, and Defect CLI orchestration now lives in
+`src/pcl/entity_handlers.py`. Parser definitions and service-layer mutation,
+transaction, event, and outbox ownership remain unchanged.
+
+## Revision
+
+- Implementation commit: `d7d38ea` (`refactor: extract entity command handlers`)
+- `src/pcl/cli.py`: 4,319 -> 3,836 lines
+- Direct handler characterization: `tests/test_entity_handlers.py`
+
+## Verification
+
+- Baseline before movement: 1,160 passed, 1 skipped.
+- Targeted entity/CLI/lifecycle tests: 67 passed.
+- Skill examples and distribution tests: 32 passed.
+- Full regression after movement: 1,164 passed, 1 skipped in 230.78s.
+- `ruff check .`: passed.
+- Source-checkout doctor: passed with zero findings.
+- Strict validation: passed with no errors; the same three active and 26
+  historical pre-existing warnings remain.
+- Render: passed.
+- `git diff --check`: passed.
+
+## Boundary review
+
+- No command, flag, help, JSON/text output, typed error, or exit-code change.
+- No dependency, schema, database migration, provider, telemetry, or external
+  write.
+- Unrelated `.claude/**`, `.playwright-cli/`, `.project-loop/*.lock`, and
+  `.work/` paths were preserved and excluded from the commit.

project_loop_harness-0.5.4/docs/evidence/0205-cli-stage3-execution-evidence-handlers.md

@@ -0,0 +1,34 @@
+# 0205 CLI Stage 3 execution and Evidence handlers evidence
+
+## Result
+
+Workflow/Loop, Jobs, Agent, Evidence, Verification, Decision, Escalation, and
+checkpoint CLI orchestration now lives in bounded execution and governance
+handler modules. Existing services retain mutation, transaction, event,
+outbox, artifact, and human-gate ownership.
+
+## Revision
+
+- Implementation commit: `dd9a327`
+- `src/pcl/cli.py`: 3,836 -> 3,111 lines
+- `src/pcl/execution_handlers.py`: 453 lines
+- `src/pcl/governance_handlers.py`: 332 lines
+- Direct characterization: `tests/test_execution_handlers.py`
+
+## Verification
+
+- Targeted execution/Evidence/governance tests: 162 passed.
+- Skill examples and distribution tests: 32 passed.
+- Full regression: 1,169 passed, 1 skipped in 224.22s.
+- `ruff check .`: passed.
+- Source-checkout doctor: passed with zero findings.
+- Strict validation: passed with no errors and the unchanged pre-existing
+  warning set (three active, 26 historical).
+- Render and `git diff --check`: passed.
+
+## Boundary review
+
+- No CLI contract, typed error, exit code, event/outbox, Evidence, or human
+  authorization behavior changed.
+- No dependency, schema, migration, provider, telemetry, or external write.
+- Unrelated dirty paths were preserved and excluded from the commit.

project_loop_harness-0.5.4/docs/evidence/0206-cli-stage3-control-profile-handlers.md

@@ -0,0 +1,35 @@
+# 0206 CLI Stage 3 control and Profile handlers evidence
+
+## Result
+
+Profile/contracts, project control, context/code-index, and planning/report
+CLI orchestration now lives in four bounded handler modules. `cli.py` retains
+parser construction, global option normalization, handler routing, and shared
+top-level error translation.
+
+## Revision
+
+- Implementation commit: `d2e4e32`
+- `src/pcl/cli.py`: 3,111 -> 1,893 lines
+- New handlers: 1,335 lines across four modules
+- Direct characterization: `tests/test_control_handlers.py`
+
+## Verification
+
+- Targeted compatibility tests: 516 passed.
+- Direct handler characterization: 4 passed.
+- Full regression: 1,173 passed, 1 skipped in 308.98s.
+- `ruff check .`: passed.
+- Source-checkout doctor: passed with zero findings.
+- Strict validation: passed with no errors and the unchanged pre-existing
+  warning set (three active, 26 historical).
+- CLI help, render, and `git diff --check`: passed.
+
+## Boundary review
+
+- Existing JSON/text/error/exit-code behavior and dry-run/human gates remain
+  unchanged.
+- Existing `pcl.cli` monkeypatch seams for time and audit failure tests remain
+  injectable through the thin router.
+- No dependency, schema, migration, provider, telemetry, or external write.
+- Unrelated dirty paths were preserved and excluded from the commit.

project_loop_harness-0.5.4/docs/evidence/0207-command-service-responsibility-split.md

@@ -0,0 +1,36 @@
+# 0207 Command service responsibility split evidence
+
+## Result
+
+`pcl.commands` is now a 54-line compatibility facade over domain operations,
+next-action routing, and finish planning. Existing public imports remain valid,
+while each responsibility has an explicit one-way dependency boundary.
+
+## Revision
+
+- Implementation commit: `d515542`
+- `src/pcl/commands.py`: 2,444 -> 54 lines
+- `src/pcl/command_domain.py`: 472 lines
+- `src/pcl/action_routing.py`: 1,785 lines
+- `src/pcl/finish_planning.py`: 214 lines
+- Direct facade characterization: `tests/test_commands_facade.py`
+
+## Verification
+
+- Targeted domain/routing/finish/dashboard tests: 156 passed.
+- Final focused facade/next/finish tests: 48 passed.
+- Full regression: 1,175 passed, 1 skipped in 324.31s.
+- `ruff check .`: passed.
+- Source-checkout doctor: passed with zero findings.
+- Strict validation: passed with no errors and the unchanged pre-existing
+  warning set (three active, 26 historical).
+- CLI help, render, and `git diff --check`: passed.
+
+## Boundary review
+
+- Public constants, call signatures, query ordering, transactions, event
+  payloads, routing priority, serialized actions, and finish plans are unchanged.
+- `start.py`, `renderer.py`, `finish_execution.py`, handlers, and direct test
+  imports continue to use `pcl.commands` without changes.
+- No dependency, schema, migration, provider, telemetry, or external write.
+- Unrelated dirty paths were preserved and excluded from the commit.

project_loop_harness-0.5.4/docs/evidence/0208-cli-stage4-parser-construction.md

@@ -0,0 +1,37 @@
+# 0208 CLI Stage 4 parser construction evidence
+
+## Result
+
+Parser construction now uses seven command-family builders behind the stable
+`pcl.parser.build_parser` facade, which remains re-exported from `pcl.cli`.
+Registration order and every existing argparse contract remain unchanged.
+
+## Revision
+
+- Implementation commit: `1a78d7c`
+- `src/pcl/cli.py`: 1,893 -> 231 lines
+- Parser facade/common code: 36 lines
+- Seven family builders: 1,628 lines
+- Direct parser characterization: `tests/test_parser_builders.py`
+
+## Verification
+
+- Existing help/baseline/Skill/parser tests: 111 passed.
+- Full command-order, top-level help, and representative-default tests: passed.
+- Wheel and sdist distribution tests: 4 passed.
+- Full regression: 1,178 passed, 1 skipped in 294.91s.
+- `ruff check .`: passed.
+- Source-checkout doctor: passed with zero findings.
+- Strict validation: passed with no errors and the unchanged pre-existing
+  warning set (three active, 26 historical).
+- CLI help, render, and `git diff --check`: passed.
+
+## Boundary review
+
+- One temporary retrieval-fixture regression was traced to an unnecessary
+  direct import in the new parser test; the test was decoupled and the failing
+  fixture plus full suite passed afterward.
+- Commands, flags, defaults, choices, aliases, help order/text, parser errors,
+  source entry point, wheel entry point, and sdist content remain compatible.
+- No dependency, schema, migration, provider, telemetry, or external write.
+- Unrelated dirty paths were preserved and excluded from the commit.

project_loop_harness-0.5.4/docs/evidence/0209-refactoring-integrated-verification.md

@@ -0,0 +1,55 @@
+# 0209 Refactoring integrated verification evidence
+
+## Result
+
+The frozen four-stage CLI and command-service refactor is complete. All
+orchestration, service, parser, public import, source, wheel, and sdist
+compatibility gates passed without intentional behavior changes.
+
+## Scope and revisions
+
+- Baseline: `1ea685a` (`cli.py` 4,319 lines; `commands.py` 2,444 lines).
+- Planning: `dd7fc7c`.
+- Entity handlers: `d7d38ea`; Evidence E-0551.
+- Execution/governance handlers: `dd9a327`; Evidence E-0552.
+- Control/Profile/context/planning handlers: `d2e4e32`; Evidence E-0553.
+- Command services and compatibility facade: `d515542`; Evidence E-0554.
+- Parser family builders: `1a78d7c`; Evidence E-0555.
+- Stage closeout documentation commits: `e0fbc79`, `318126d`, `ddcf2e9`,
+  `2c614ed`, and `38b505a`.
+
+## Final architecture
+
+- `src/pcl/cli.py`: 231-line global-option, routing, and error facade.
+- `src/pcl/commands.py`: 54-line stable import facade.
+- Command handlers: read, entity, execution, governance, profile, control,
+  context, and planning families.
+- Services: domain operations, next-action routing, and finish planning.
+- Parser: one facade, one shared helper, and seven ordered family builders.
+
+## Final verification
+
+- Full regression: 1,178 passed, 1 skipped in 294.91s.
+- New characterization coverage: 18 tests across handler, command-facade, and
+  parser-family boundaries.
+- Baseline/help/Skill parser tests and all top-level command help smoke passed.
+- Wheel install and sdist content/runtime smoke passed.
+- Ruff and `git diff --check` passed.
+- Source-checkout doctor passed with zero findings.
+- Strict validation passed with no errors; the pre-existing warning set stayed
+  at three active and 26 historical findings.
+- Render passed. Audit check confirmed 1,903 SQLite events, JSONL events, and
+  delivered outbox rows with no pending/failed rows. It returned exit 6 for 55
+  pre-existing human-review Evidence anomalies (E-0018 through E-0508); none
+  belongs to the refactor Evidence E-0551 through E-0555.
+
+## Boundary and residual state
+
+- No command, flag, help, output, exit code, typed error, transaction, event,
+  outbox, Evidence, human-gate, schema, dependency, or generated-artifact
+  contract changed intentionally.
+- No publish, push, provider call, telemetry, or production write occurred.
+- Existing `.claude`, `.playwright-cli`, `.work`, and Project Loop lock files
+  remain unrelated and uncommitted.
+- Pre-existing E-0018 and E-0182 integrity warnings remain outside this
+  refactor and were not modified.