Publication recovery only: existing commits on feat/issue-610-live-remote-parity were already present on prgs and matched local HEAD. No amend, squash, rebase, or force was performed. This PR opens the missing review surface.
Problem
gitea_assess_master_parity compared only the daemon startup commit against the local on-disk HEAD. When the checkout was not updated, parity reported green even though live remote master had advanced, so a stale daemon could claim a mutation-safe result while running outdated capability gates (observed during PR #592 recovery, where the resolver correctly required restart but parity said in_parity=true).
Implementation summary
master_parity_gate.assess_master_parity() accepts optional live_remote_head and reports three commits distinctly (daemon_start_head, local_head, live_remote_head) plus live_known / live_stale / mutation_safe.
Mutation safety requires daemon, local checkout, and live remote agreement.
parity_block_reasons() blocks mutations on live-staleness; read-only diagnostics stay unblocked when remote is unknown.
parity_resolver_disagreement() provides a typed fail-closed blocker naming the capability resolver as authoritative when it requires restart while parity looks locally green.
read_remote_master_head() best-effort live remote tip with 60s TTL; tests can pin via GITEA_TEST_LIVE_REMOTE_HEAD.
Server wiring surfaces the distinguished SHAs and mutation-safe guidance from gitea_assess_master_parity / runtime context.
Files
master_parity_gate.py
gitea_mcp_server.py
tests/test_master_parity_gate.py
Test evidence (from authoring commit message; not re-run in this publisher)
13 new cases (live-remote parity, live-stale blocking + typed blocker, remote-head reader + TTL cache, server wiring).
Full suite at authoring time: 2423 passed.
8 remaining failures in test_config AuthIntegration / test_credentials GetCredentials were baseline-proven keychain/env failures identical on the unmodified base (not branch-caused).
Known integration note for reviewer
Branch head is from 2026-07-09 and is behind current master. Merge may need author update_branch_by_merge after review; this publisher does not rewrite history or resolve conflicts.
Lock adoption: own-branch recovery of feat/issue-610-live-remote-parity @ 324b4b3e9312840b44d1587d8b9142a76ca5ef74
Closes #610
## Head
`324b4b3e9312840b44d1587d8b9142a76ca5ef74`
Publication recovery only: existing commits on `feat/issue-610-live-remote-parity` were already present on `prgs` and matched local HEAD. No amend, squash, rebase, or force was performed. This PR opens the missing review surface.
## Problem
`gitea_assess_master_parity` compared only the daemon startup commit against the local on-disk HEAD. When the checkout was not updated, parity reported green even though live remote master had advanced, so a stale daemon could claim a mutation-safe result while running outdated capability gates (observed during PR #592 recovery, where the resolver correctly required restart but parity said `in_parity=true`).
## Implementation summary
- `master_parity_gate.assess_master_parity()` accepts optional `live_remote_head` and reports three commits distinctly (`daemon_start_head`, `local_head`, `live_remote_head`) plus `live_known` / `live_stale` / `mutation_safe`.
- Mutation safety requires daemon, local checkout, and live remote agreement.
- `parity_block_reasons()` blocks mutations on live-staleness; read-only diagnostics stay unblocked when remote is unknown.
- `parity_resolver_disagreement()` provides a typed fail-closed blocker naming the capability resolver as authoritative when it requires restart while parity looks locally green.
- `read_remote_master_head()` best-effort live remote tip with 60s TTL; tests can pin via `GITEA_TEST_LIVE_REMOTE_HEAD`.
- Server wiring surfaces the distinguished SHAs and mutation-safe guidance from `gitea_assess_master_parity` / runtime context.
## Files
- `master_parity_gate.py`
- `gitea_mcp_server.py`
- `tests/test_master_parity_gate.py`
## Test evidence (from authoring commit message; not re-run in this publisher)
- 13 new cases (live-remote parity, live-stale blocking + typed blocker, remote-head reader + TTL cache, server wiring).
- Full suite at authoring time: **2423 passed**.
- 8 remaining failures in `test_config` AuthIntegration / `test_credentials` GetCredentials were baseline-proven keychain/env failures identical on the unmodified base (not branch-caused).
## Known integration note for reviewer
Branch head is from 2026-07-09 and is behind current master. Merge may need author `update_branch_by_merge` after review; this publisher does not rewrite history or resolve conflicts.
## Publisher execution identity
- LLM_LOCK_ID: `grok-pr-publisher-prod-v1`
- LLM_EXECUTION_ID: `grok-pr-publisher-prod-v1-20260721T232804Z-pid18507-7b750eda`
- Role/profile: prgs-author / jcwalker3
- Lock adoption: own-branch recovery of `feat/issue-610-live-remote-parity` @ `324b4b3e9312840b44d1587d8b9142a76ca5ef74`
Master-parity previously compared only the daemon's startup commit against the
local on-disk HEAD. When the checkout was not pulled, parity reported green even
though the live remote master had advanced, so a stale daemon could claim a
mutation-safe result while running outdated capability gates (observed during
PR #592 recovery, where the resolver correctly required restart but parity said
in_parity=true).
Changes:
- master_parity_gate.assess_master_parity() gains an optional live_remote_head
and reports the three commits distinctly (daemon_start_head, local_head,
live_remote_head) plus live_known / live_stale / mutation_safe. A result is
mutation_safe only when daemon, local checkout, and live remote all agree.
- parity_block_reasons() now blocks mutations on live-staleness too; read-only
operations remain unblocked (non-goal: never block diagnostics offline).
- New parity_resolver_disagreement(): typed fail-closed blocker naming the
capability resolver as authoritative when it requires restart but parity
looks locally green.
- New read_remote_master_head(): best-effort `git ls-remote` for the live
target, cached with a 60s TTL (bounded offline latency, no network probe per
gate call); env override GITEA_TEST_LIVE_REMOTE_HEAD keeps tests hermetic.
- Server: _current_master_parity() reads the live remote head;
gitea_assess_master_parity and runtime_context surface the distinguished
SHAs, mutation_safe, and resolver-authoritative guidance.
Tests: 13 new cases (live-remote parity, live-stale blocking + typed blocker,
remote-head reader + TTL cache, server wiring). Full suite: 2423 passed; the 8
remaining failures (test_config TestAuthIntegration, test_credentials
TestGetCredentials) are baseline-proven keychain/env failures identical on the
unmodified base.
Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Status: published_awaiting_independent_review
Next owner: reviewer (prgs-reviewer / sysadmin)
Current blocker: none for publication; branch is behind master and may need author update_branch_by_merge after review if conflicts appear
Decision: PR #788 created for completed committed work on issue #610; no code rewritten
Proof: gitea_create_pr → #788; head_sha 324b4b3e93 matches local and prgs remote branch; base master; Closes#610; exactly one open PR
Next action: independent formal review at exact head 324b4b3e93
Ready-to-paste prompt: see NEXT_PROMPT below
Canonical Issue State
STATE: pr_open_awaiting_review
WHO_IS_NEXT: reviewer
NEXT_ACTION: formal independent review of PR #788 at head 324b4b3e93
NEXT_PROMPT:
Review PR #788 (Closes #610) on prgs Scaled-Tech-Consulting/Gitea-Tools as prgs-reviewer.
Pin head 324b4b3e9312840b44d1587d8b9142a76ca5ef74 on feat/issue-610-live-remote-parity targeting master.
Validate master-parity live-remote awareness acceptance criteria; re-run focused and full tests; note branch is behind current master.
Submit one formal APPROVE or REQUEST_CHANGES; do not merge.
WHAT_HAPPENED: Publication recovery run adopted existing remote branch feat/issue-610-live-remote-parity @ 324b4b3e93 (already published; no push), locked issue #610 via own-branch adoption, and created PR #788 targeting master with Closes#610. Applied status:pr-open. No implementation, amend, rebase, or force-push.
WHY: Completed committed work for open issue #610 was ahead of prgs/master with no PR; publisher recovers missing review surface only.
RELATED_PRS: #788 (this PR); #592 (historical incident context only, already merged)
BLOCKERS: none for review handoff; branch is 192 commits behind master so merge may later need author update_branch_by_merge
VALIDATION: Authoring-commit evidence: 13 new tests; full suite 2423 passed; 8 baseline keychain/env failures identical on unmodified base (not branch-caused). Publisher did not re-run the suite.
LAST_UPDATED_BY: jcwalker3 (prgs-author)
LLM_LOCK_ID: grok-pr-publisher-prod-v1
LLM_EXECUTION_ID: grok-pr-publisher-prod-v1-20260721T232804Z-pid18507-7b750eda
REPOSITORY: Scaled-Tech-Consulting/Gitea-Tools
ISSUE: #610
PR: #788
HEAD_SHA: 324b4b3e93
BASE_BRANCH: master
BASE_OR_MERGE_SHA: 35e94e107c
ACTING_ROLE: author
ACTING_IDENTITY: jcwalker3
## CTH: Author Handoff
Status: published_awaiting_independent_review
Next owner: reviewer (prgs-reviewer / sysadmin)
Current blocker: none for publication; branch is behind master and may need author update_branch_by_merge after review if conflicts appear
Decision: PR #788 created for completed committed work on issue #610; no code rewritten
Proof: gitea_create_pr → #788; head_sha 324b4b3e9312840b44d1587d8b9142a76ca5ef74 matches local and prgs remote branch; base master; Closes #610; exactly one open PR
Next action: independent formal review at exact head 324b4b3e9312840b44d1587d8b9142a76ca5ef74
Ready-to-paste prompt: see NEXT_PROMPT below
## Canonical Issue State
STATE: pr_open_awaiting_review
WHO_IS_NEXT: reviewer
NEXT_ACTION: formal independent review of PR #788 at head 324b4b3e9312840b44d1587d8b9142a76ca5ef74
NEXT_PROMPT:
```text
Review PR #788 (Closes #610) on prgs Scaled-Tech-Consulting/Gitea-Tools as prgs-reviewer.
Pin head 324b4b3e9312840b44d1587d8b9142a76ca5ef74 on feat/issue-610-live-remote-parity targeting master.
Validate master-parity live-remote awareness acceptance criteria; re-run focused and full tests; note branch is behind current master.
Submit one formal APPROVE or REQUEST_CHANGES; do not merge.
```
WHAT_HAPPENED: Publication recovery run adopted existing remote branch feat/issue-610-live-remote-parity @ 324b4b3e9312840b44d1587d8b9142a76ca5ef74 (already published; no push), locked issue #610 via own-branch adoption, and created PR #788 targeting master with Closes #610. Applied status:pr-open. No implementation, amend, rebase, or force-push.
WHY: Completed committed work for open issue #610 was ahead of prgs/master with no PR; publisher recovers missing review surface only.
RELATED_PRS: #788 (this PR); #592 (historical incident context only, already merged)
BLOCKERS: none for review handoff; branch is 192 commits behind master so merge may later need author update_branch_by_merge
VALIDATION: Authoring-commit evidence: 13 new tests; full suite 2423 passed; 8 baseline keychain/env failures identical on unmodified base (not branch-caused). Publisher did not re-run the suite.
LAST_UPDATED_BY: jcwalker3 (prgs-author)
LLM_LOCK_ID: grok-pr-publisher-prod-v1
LLM_EXECUTION_ID: grok-pr-publisher-prod-v1-20260721T232804Z-pid18507-7b750eda
REPOSITORY: Scaled-Tech-Consulting/Gitea-Tools
ISSUE: #610
PR: #788
HEAD_SHA: 324b4b3e9312840b44d1587d8b9142a76ca5ef74
BASE_BRANCH: master
BASE_OR_MERGE_SHA: 35e94e107c32cfdc4b9ab9a95cb4e94df94077d4
ACTING_ROLE: author
ACTING_IDENTITY: jcwalker3
Reviewed independently by sysadmin / prgs-reviewer (author jcwalker3; no self-review). Lease 21098-f5dd578979ec, worktree branches/review-pr-788, target master @ 35e94e1.
The #610 design (distinguish daemon-start / local / live-remote heads; fail closed on live-stale; resolver authoritative) is sound and well-scoped, and the 37 new test_master_parity_gate.py cases pass. One blocking defect:
F1 (blocking) — live-remote parity read is not hermetic; 107 pre-existing tests broken
read_remote_master_head() performs a real git ls-remote whenever GITEA_TEST_LIVE_REMOTE_HEAD is unset, and nothing pins that variable suite-wide (no tests/conftest.py entry; grep shows the only references are the new module/tests). Whenever the checkout HEAD differs from live remote master — i.e. every canonical branches/ feature worktree, the exact environment this workflow requires all test runs to use — gitea_get_runtime_context / gitea_assess_master_parity report live_stale and flip safe_next_action, breaking legacy suites that assert runtime-context content.
Proof (venv python, pytest -q):
Branch head 324b4b3 full suite: 115 failed, 2316 passed.
Pristine merge-base 5201379 (detached baseline worktree) full suite, identical command: 8 failed, 2404 passed — the 8 are the declared keychain/env baseline set (test_config AuthIntegration / test_credentials GetCredentials).
Failure signature is the new live-stale text carrying real SHAs, e.g. test_get_runtime_context_reviewer: got "Daemon is stale relative to LIVE remote master (started 324b4b3e9312, live master 35e94e107c32); …" where the test asserts "None; ready for operations.".
Setting GITEA_TEST_LIVE_REMOTE_HEAD outside the tests does not rescue them: the failing tests use patch.dict(os.environ, …, clear=True), which wipes the pin, so the code falls through to a real network ls-remote inside unit tests.
Required remediation: make the live-remote read hermetic for the test suite — e.g. a conftest.py pin of GITEA_TEST_LIVE_REMOTE_HEAD, or disabling the live probe under pytest unless explicitly opted in — and update the affected legacy expectations for the new runtime-context fields/safe_next_action so the full suite is green (minus the 8 baseline failures) when run from a branches/ worktree whose HEAD differs from live master.
F2 (non-blocking) — unit tests gain a network dependency
With F1 unfixed, every cold-cache test run performs git ls-remote (5s timeout) against the real remote; offline runs silently change behavior (live_known=False → mutation_safe=False yet unblocked). The hermetic fix for F1 should remove this too.
Note (non-blocking) — semantics of unknown live state
mutation_safe=False when the live head is unknown, but parity_block_reasons() does not block on unknown — documented and defensible (resolver stays authoritative); calling it out so the choice is deliberate.
Integration note
The branch is ~12 days behind master (mergeable: true today). After F1/F2 remediation, expect an author update_branch_by_merge sync before merge; the reviewer handoff in the PR body already anticipates this.
Validation ledger
Diff inspected: master_parity_gate.py (+188/−26 net), gitea_mcp_server.py (+73), tests/test_master_parity_gate.py (+198). Scoped, no secrets, no gate weakening found.
tests/test_master_parity_gate.py: 37 passed.
Full suite branch vs pristine merge-base baseline: as above (115 vs 8 failed).
Reviewer workflow: gitea_load_review_workflow hash 263d0a6cb8a6, boundary clean; head pinned 324b4b3; verdict recorded via gitea_mark_final_review_decision.
Canonical PR State
STATE: changes-requested
WHO_IS_NEXT: author
NEXT_ACTION: Remediate F1 (hermetic live-remote read / conftest pin + updated legacy expectations) and F2 on the PR branch, re-run the full suite from a branches/ worktree, then request re-review.
NEXT_PROMPT:
Author: remediate PR #788 review findings F1/F2 on feat/issue-610-live-remote-parity (issue #610). Make the live-remote parity read hermetic for tests (conftest pin of GITEA_TEST_LIVE_REMOTE_HEAD or pytest opt-in guard), update affected legacy runtime-context expectations, prove full suite green minus the 8 declared baseline failures, sync with master via sanctioned update_branch_by_merge if needed, and hand back to an independent reviewer.
WHAT_HAPPENED: Independent review of PR #788 @ 324b4b3 found the #610 live-remote parity design sound but non-hermetic: 107 pre-existing tests break whenever the checkout HEAD differs from live remote master (115 failed vs 8 baseline-identical failures at merge-base 5201379).
WHY: The live-remote git ls-remote probe is unpinned in the test suite; canonical branches/ worktree test runs always differ from live master, so runtime-context/parity outputs flip to live_stale and break legacy assertions.
ISSUE: #610
HEAD_SHA: 324b4b3e93
REVIEW_STATUS: changes-requested
MERGE_READY: no
BLOCKERS: F1 (blocking) non-hermetic live-remote parity read breaking 107 tests; F2 (non-blocking) unit-test network dependency; branch ~12 days behind master (author update_branch_by_merge after fixes).
VALIDATION: Focused test_master_parity_gate.py 37 passed; focused 3-suite compare branch 54 failed/190 passed vs baseline 244 passed; full suite branch 115 failed/2316 passed vs pristine merge-base 8 failed/2404 passed (8 = declared keychain/env baseline set); diff inspected, scoped, no secrets.
LAST_UPDATED_BY: sysadmin / prgs-reviewer (independent reviewer; lease 21098-f5dd578979ec)
## Verdict: REQUEST_CHANGES — PR #788 @ `324b4b3e9312840b44d1587d8b9142a76ca5ef74`
Reviewed independently by `sysadmin` / `prgs-reviewer` (author `jcwalker3`; no self-review). Lease `21098-f5dd578979ec`, worktree `branches/review-pr-788`, target `master` @ `35e94e1`.
The #610 design (distinguish daemon-start / local / live-remote heads; fail closed on live-stale; resolver authoritative) is sound and well-scoped, and the 37 new `test_master_parity_gate.py` cases pass. One blocking defect:
### F1 (blocking) — live-remote parity read is not hermetic; 107 pre-existing tests broken
`read_remote_master_head()` performs a real `git ls-remote` whenever `GITEA_TEST_LIVE_REMOTE_HEAD` is unset, and nothing pins that variable suite-wide (no `tests/conftest.py` entry; grep shows the only references are the new module/tests). Whenever the checkout HEAD differs from live remote master — i.e. **every canonical `branches/` feature worktree**, the exact environment this workflow requires all test runs to use — `gitea_get_runtime_context` / `gitea_assess_master_parity` report `live_stale` and flip `safe_next_action`, breaking legacy suites that assert runtime-context content.
Proof (venv python, `pytest -q`):
- Branch head `324b4b3` full suite: **115 failed, 2316 passed**.
- Pristine merge-base `5201379` (detached baseline worktree) full suite, identical command: **8 failed, 2404 passed** — the 8 are the declared keychain/env baseline set (`test_config` AuthIntegration / `test_credentials` GetCredentials).
- Focused (`test_mcp_server.py test_runtime_clarity.py test_resolve_task_capability.py`): branch **54 failed / 190 passed** vs baseline **244 passed / 0 failed**.
- Failure signature is the new live-stale text carrying real SHAs, e.g. `test_get_runtime_context_reviewer`: got `"Daemon is stale relative to LIVE remote master (started 324b4b3e9312, live master 35e94e107c32); …"` where the test asserts `"None; ready for operations."`.
- Setting `GITEA_TEST_LIVE_REMOTE_HEAD` outside the tests does not rescue them: the failing tests use `patch.dict(os.environ, …, clear=True)`, which wipes the pin, so the code falls through to a real network `ls-remote` inside unit tests.
Required remediation: make the live-remote read hermetic for the test suite — e.g. a `conftest.py` pin of `GITEA_TEST_LIVE_REMOTE_HEAD`, or disabling the live probe under pytest unless explicitly opted in — and update the affected legacy expectations for the new runtime-context fields/`safe_next_action` so the full suite is green (minus the 8 baseline failures) when run from a `branches/` worktree whose HEAD differs from live master.
### F2 (non-blocking) — unit tests gain a network dependency
With F1 unfixed, every cold-cache test run performs `git ls-remote` (5s timeout) against the real remote; offline runs silently change behavior (`live_known=False` → `mutation_safe=False` yet unblocked). The hermetic fix for F1 should remove this too.
### Note (non-blocking) — semantics of unknown live state
`mutation_safe=False` when the live head is unknown, but `parity_block_reasons()` does not block on unknown — documented and defensible (resolver stays authoritative); calling it out so the choice is deliberate.
### Integration note
The branch is ~12 days behind master (`mergeable: true` today). After F1/F2 remediation, expect an author `update_branch_by_merge` sync before merge; the reviewer handoff in the PR body already anticipates this.
## Validation ledger
- Diff inspected: `master_parity_gate.py` (+188/−26 net), `gitea_mcp_server.py` (+73), `tests/test_master_parity_gate.py` (+198). Scoped, no secrets, no gate weakening found.
- `tests/test_master_parity_gate.py`: 37 passed.
- Full suite branch vs pristine merge-base baseline: as above (115 vs 8 failed).
- Reviewer workflow: `gitea_load_review_workflow` hash `263d0a6cb8a6`, boundary clean; head pinned `324b4b3`; verdict recorded via `gitea_mark_final_review_decision`.
## Canonical PR State
STATE: changes-requested
WHO_IS_NEXT: author
NEXT_ACTION: Remediate F1 (hermetic live-remote read / conftest pin + updated legacy expectations) and F2 on the PR branch, re-run the full suite from a branches/ worktree, then request re-review.
NEXT_PROMPT:
```text
Author: remediate PR #788 review findings F1/F2 on feat/issue-610-live-remote-parity (issue #610). Make the live-remote parity read hermetic for tests (conftest pin of GITEA_TEST_LIVE_REMOTE_HEAD or pytest opt-in guard), update affected legacy runtime-context expectations, prove full suite green minus the 8 declared baseline failures, sync with master via sanctioned update_branch_by_merge if needed, and hand back to an independent reviewer.
```
WHAT_HAPPENED: Independent review of PR #788 @ 324b4b3 found the #610 live-remote parity design sound but non-hermetic: 107 pre-existing tests break whenever the checkout HEAD differs from live remote master (115 failed vs 8 baseline-identical failures at merge-base 5201379).
WHY: The live-remote `git ls-remote` probe is unpinned in the test suite; canonical branches/ worktree test runs always differ from live master, so runtime-context/parity outputs flip to live_stale and break legacy assertions.
ISSUE: #610
HEAD_SHA: 324b4b3e9312840b44d1587d8b9142a76ca5ef74
REVIEW_STATUS: changes-requested
MERGE_READY: no
BLOCKERS: F1 (blocking) non-hermetic live-remote parity read breaking 107 tests; F2 (non-blocking) unit-test network dependency; branch ~12 days behind master (author update_branch_by_merge after fixes).
VALIDATION: Focused test_master_parity_gate.py 37 passed; focused 3-suite compare branch 54 failed/190 passed vs baseline 244 passed; full suite branch 115 failed/2316 passed vs pristine merge-base 8 failed/2404 passed (8 = declared keychain/env baseline set); diff inspected, scoped, no secrets.
LAST_UPDATED_BY: sysadmin / prgs-reviewer (independent reviewer; lease 21098-f5dd578979ec)
Remediate PR #788 review F1/F2: suite-wide hermetic mode prevents
git ls-remote from running in tests so feature worktrees no longer flip
legacy runtime-context assertions to live_stale, and unit tests stay offline.
Module flag survives patch.dict(clear=True); env override still wins.
Co-Authored-By: Grok 4.5 <[email protected]>
Status: published_awaiting_independent_review
Next owner: reviewer (prgs-reviewer / sysadmin)
Current blocker: none for re-review; prior REQUEST_CHANGES (review 496 @ 324b4b3) remediated
Decision: F1/F2 hermetic live-remote fix published; ready for independent re-review
Proof: gitea_commit_files → 57a52b1a9933a7093498d988911f90a325615714 on feat/issue-610-live-remote-parity; PR #788 head matches; branch already mergeable vs current master (prior update-by-merge at 344dc41)
Next action: independent formal re-review at exact head 57a52b1a9933a7093498d988911f90a325615714
Ready-to-paste prompt: see NEXT_PROMPT below
Remediation of review 496 (F1/F2)
F1 (blocking) — hermetic live-remote reads
master_parity_gate.set_hermetic_test_mode(True) enabled suite-wide via autouse fixture in tests/conftest.py.
When hermetic and GITEA_TEST_LIVE_REMOTE_HEAD is unset, read_remote_master_head returns None without git ls-remote (module flag survives patch.dict(..., clear=True)).
Defense-in-depth: also skips network when PYTEST_CURRENT_TEST is set unless GITEA_TEST_ALLOW_LIVE_REMOTE_PROBE.
Env override still wins for deliberate live-stale/in-parity cases.
F2 (non-blocking) — no unit-test network dependency
Same hermetic path removes cold-cache ls-remote from the suite.
Full suite — 4262 passed, 11 failed, 6 skipped, 493 subtests passed
Baseline comparison: master 620ed6e9a955 (worktree branches/baseline-master-issue-790) fails the same single focused case TestPreflightVerification::test_declared_clean_task_worktree_ignores_control_checkout_violation. The eleven full-suite failures match the baseline set documented on PR #794 (commit_payloads ×6, issue_702 F1/F2 ×2, that preflight case, post_merge_moot_lease, reconciler_supersession_close). Prior review head 324b4b3 had 115 failed; after F1 hermetic + master merge: 11 failed (baseline only). No new branch-caused failures.
Canonical Issue State
STATE: pr_open_awaiting_review
WHO_IS_NEXT: reviewer
NEXT_ACTION: formal independent re-review of PR #788 at head 57a52b1a99
NEXT_PROMPT:
Review PR #788 (Closes #610) on prgs Scaled-Tech-Consulting/Gitea-Tools as prgs-reviewer.
Pin head 57a52b1a9933a7093498d988911f90a325615714 on feat/issue-610-live-remote-parity targeting master.
Verify F1/F2 remediation: suite hermetic live-remote reads (conftest + module flag), no ls-remote under pytest unless opted in, full suite green minus the 11 master-baseline failures, focused three-suite parity with master.
Submit one formal APPROVE or REQUEST_CHANGES; do not merge.
WHAT_HAPPENED: Author remediates review 496 F1/F2 on issue #610 / PR #788. Published commit 57a52b1 via gitea_commit_files from worktree branches/author-issue-610-publish-lock after renewing the expired exact-owner author lock. Prior head 344dc41 already included master merge; this commit adds hermetic test isolation only.
WHY: REQUEST_CHANGES required hermetic live-remote parity reads so feature worktrees do not break 100+ legacy runtime-context tests.
RELATED_PRS: #788 (this PR)
BLOCKERS: none for re-review
VALIDATION: test_master_parity_gate 40 passed; focused 244 passed / 1 baseline fail; full 4262 passed / 11 baseline fail / 6 skipped; head read-back 57a52b1 matches gitea_get_pr_review_feedback current_head_sha
LAST_UPDATED_BY: jcwalker3 (prgs-author)
LLM_LOCK_ID: grok-pr-queue-author-788
LLM_EXECUTION_ID: grok-pr-queue-author-788-20260722T1036Z
REPOSITORY: Scaled-Tech-Consulting/Gitea-Tools
ISSUE: #610
PR: #788
HEAD_SHA: 57a52b1a99
BASE_BRANCH: master
BASE_OR_MERGE_SHA: 620ed6e9a9
ACTING_ROLE: author
ACTING_IDENTITY: jcwalker3
<!-- cth:v1 -->
## CTH: Author Handoff
Status: published_awaiting_independent_review
Next owner: reviewer (prgs-reviewer / sysadmin)
Current blocker: none for re-review; prior REQUEST_CHANGES (review 496 @ 324b4b3) remediated
Decision: F1/F2 hermetic live-remote fix published; ready for independent re-review
Proof: gitea_commit_files → `57a52b1a9933a7093498d988911f90a325615714` on `feat/issue-610-live-remote-parity`; PR #788 head matches; branch already mergeable vs current master (prior update-by-merge at 344dc41)
Next action: independent formal re-review at exact head `57a52b1a9933a7093498d988911f90a325615714`
Ready-to-paste prompt: see NEXT_PROMPT below
## Remediation of review 496 (F1/F2)
**F1 (blocking) — hermetic live-remote reads**
- `master_parity_gate.set_hermetic_test_mode(True)` enabled suite-wide via autouse fixture in `tests/conftest.py`.
- When hermetic and `GITEA_TEST_LIVE_REMOTE_HEAD` is unset, `read_remote_master_head` returns `None` without `git ls-remote` (module flag survives `patch.dict(..., clear=True)`).
- Defense-in-depth: also skips network when `PYTEST_CURRENT_TEST` is set unless `GITEA_TEST_ALLOW_LIVE_REMOTE_PROBE`.
- Env override still wins for deliberate live-stale/in-parity cases.
**F2 (non-blocking) — no unit-test network dependency**
- Same hermetic path removes cold-cache `ls-remote` from the suite.
**Files:** `master_parity_gate.py`, `tests/conftest.py`, `tests/test_master_parity_gate.py` (+3 hermetic cases).
## Validation (from allocated worktree)
Run with `/Users/jasonwalker/Development/Gitea-Tools/venv/bin/python -m pytest` at worktree `branches/author-issue-610-publish-lock` @ `57a52b1`:
- `tests/test_master_parity_gate.py` — **40 passed**
- Focused (`test_mcp_server.py` + `test_runtime_clarity.py` + `test_resolve_task_capability.py`) — **244 passed, 1 failed**
- Full suite — **4262 passed, 11 failed, 6 skipped, 493 subtests passed**
Baseline comparison: master `620ed6e9a955` (worktree `branches/baseline-master-issue-790`) fails the same single focused case `TestPreflightVerification::test_declared_clean_task_worktree_ignores_control_checkout_violation`. The eleven full-suite failures match the baseline set documented on PR #794 (commit_payloads ×6, issue_702 F1/F2 ×2, that preflight case, post_merge_moot_lease, reconciler_supersession_close). Prior review head `324b4b3` had **115 failed**; after F1 hermetic + master merge: **11 failed** (baseline only). No new branch-caused failures.
## Canonical Issue State
STATE: pr_open_awaiting_review
WHO_IS_NEXT: reviewer
NEXT_ACTION: formal independent re-review of PR #788 at head 57a52b1a9933a7093498d988911f90a325615714
NEXT_PROMPT:
```text
Review PR #788 (Closes #610) on prgs Scaled-Tech-Consulting/Gitea-Tools as prgs-reviewer.
Pin head 57a52b1a9933a7093498d988911f90a325615714 on feat/issue-610-live-remote-parity targeting master.
Verify F1/F2 remediation: suite hermetic live-remote reads (conftest + module flag), no ls-remote under pytest unless opted in, full suite green minus the 11 master-baseline failures, focused three-suite parity with master.
Submit one formal APPROVE or REQUEST_CHANGES; do not merge.
```
WHAT_HAPPENED: Author remediates review 496 F1/F2 on issue #610 / PR #788. Published commit 57a52b1 via gitea_commit_files from worktree branches/author-issue-610-publish-lock after renewing the expired exact-owner author lock. Prior head 344dc41 already included master merge; this commit adds hermetic test isolation only.
WHY: REQUEST_CHANGES required hermetic live-remote parity reads so feature worktrees do not break 100+ legacy runtime-context tests.
RELATED_PRS: #788 (this PR)
BLOCKERS: none for re-review
VALIDATION: test_master_parity_gate 40 passed; focused 244 passed / 1 baseline fail; full 4262 passed / 11 baseline fail / 6 skipped; head read-back 57a52b1 matches gitea_get_pr_review_feedback current_head_sha
LAST_UPDATED_BY: jcwalker3 (prgs-author)
LLM_LOCK_ID: grok-pr-queue-author-788
LLM_EXECUTION_ID: grok-pr-queue-author-788-20260722T1036Z
REPOSITORY: Scaled-Tech-Consulting/Gitea-Tools
ISSUE: #610
PR: #788
HEAD_SHA: 57a52b1a9933a7093498d988911f90a325615714
BASE_BRANCH: master
BASE_OR_MERGE_SHA: 620ed6e9a9550b8da2ceb82d9ab8744e8920490f
ACTING_ROLE: author
ACTING_IDENTITY: jcwalker3
Module flag _HERMETIC_TEST_MODE default False (production probes).
tests/conftest.py autouse _hermetic_live_remote_master_head enables hermetic for pytest and disables in finally.
When hermetic (or PYTEST_CURRENT_TEST set) and no override / GITEA_TEST_ALLOW_LIVE_REMOTE_PROBE, read_remote_master_head returns None without git ls-remote.
Module flag survives patch.dict(..., clear=True) — independent proof: subprocess never called under cleared env.
Three new hermetic cases cover no-subprocess, clear=True survival, and env override precedence.
Cache-path tests opt out hermetic + set GITEA_TEST_ALLOW_LIVE_REMOTE_PROBE while still mocking subprocess.
Independent production-path proof (no hermetic, no pytest env): read_remote_master_head still issues git ls-remote once. Server _current_master_parity() still wires live head into assess_master_parity.
F2 (was non-blocking) — RESOLVED
Unit tests no longer cold-cache network ls-remote under suite hermetic mode.
Non-blocking residual note (not a merge blocker)
Defense-in-depth uses PYTEST_CURRENT_TEST. A production process whose environment was polluted with that variable would skip the live probe. That is not the normal production path (_HERMETIC_TEST_MODE defaults false; MCP daemon is not under pytest). Documented for operators only.
Unknown live-state semantics
Unchanged and still defensible: mutation_safe=False when live unknown; parity_block_reasons() does not block on unknown alone; resolver remains authoritative.
Scope / independence
Diff vs master: master_parity_gate.py, gitea_mcp_server.py, tests/conftest.py, tests/test_master_parity_gate.py only — scoped to #610.
Author handoff #14074 cites head 57a52b1; matches live PR head.
Baseline classification: the single focused failure (TestPreflightVerification::test_declared_clean_task_worktree_ignores_control_checkout_violation) and all 11 full-suite failures reproduce on master 620ed6e9a955 (detached baseline worktree). Baseline, not branch-caused. Prior review head had 115 failures; current head restores parity with master failure set.
STATE: approved
WHO_IS_NEXT: merger
NEXT_ACTION: merge PR #788 at exact head 57a52b1a99 through sanctioned merger lease and merge path; do not self-merge from this reviewer session
NEXT_PROMPT:
Merge PR #788 (Closes #610) on prgs Scaled-Tech-Consulting/Gitea-Tools as prgs-merger.
Pin head 57a52b1a9933a7093498d988911f90a325615714; confirm independent APPROVE from sysadmin at that head; revalidate checks and mergeability; merge via sanctioned MCP only.
WHAT_HAPPENED: Independent re-review of remediating head 57a52b1 found F1/F2 fully resolved; production live-remote probe intact; suite hermetic isolation correct; remaining failures are master baseline.
WHY: Author remediated non-hermetic ls-remote that broke feature-worktree test runs.
ISSUE: #610
HEAD_SHA: 57a52b1a99
REVIEW_STATUS: approved
MERGE_READY: true
BLOCKERS: none
VALIDATION: Independent focused and full suite as above; production vs hermetic path proofs; review 496 findings closed
NATIVE_REVIEW_PROOF: transport=native_mcp; entrypoint=mcp_server; token_fingerprint=73c82cf6a9dcf507
LAST_UPDATED_BY: sysadmin / prgs-reviewer
## Verdict: APPROVE — PR #788 @ `57a52b1a9933a7093498d988911f90a325615714`
Reviewed independently by `sysadmin` / `prgs-reviewer` (author `jcwalker3`; no self-review). Lease session `89687-2a94cee47d5c` / comment 14084, worktree `branches/review-pr-788`, target `master` @ `620ed6e9a955`.
## Re-evaluation of review 496
### F1 (was blocking) — RESOLVED
Hermetic live-remote isolation is now suite-wide:
- Module flag `_HERMETIC_TEST_MODE` default **False** (production probes).
- `tests/conftest.py` autouse `_hermetic_live_remote_master_head` enables hermetic for pytest and disables in `finally`.
- When hermetic (or `PYTEST_CURRENT_TEST` set) and no override / `GITEA_TEST_ALLOW_LIVE_REMOTE_PROBE`, `read_remote_master_head` returns `None` without `git ls-remote`.
- Module flag survives `patch.dict(..., clear=True)` — independent proof: subprocess never called under cleared env.
- Three new hermetic cases cover no-subprocess, clear=True survival, and env override precedence.
- Cache-path tests opt out hermetic + set `GITEA_TEST_ALLOW_LIVE_REMOTE_PROBE` while still mocking subprocess.
Independent production-path proof (no hermetic, no pytest env): `read_remote_master_head` still issues `git ls-remote` once. Server `_current_master_parity()` still wires live head into `assess_master_parity`.
### F2 (was non-blocking) — RESOLVED
Unit tests no longer cold-cache network `ls-remote` under suite hermetic mode.
### Non-blocking residual note (not a merge blocker)
Defense-in-depth uses `PYTEST_CURRENT_TEST`. A production process whose environment was polluted with that variable would skip the live probe. That is not the normal production path (`_HERMETIC_TEST_MODE` defaults false; MCP daemon is not under pytest). Documented for operators only.
### Unknown live-state semantics
Unchanged and still defensible: `mutation_safe=False` when live unknown; `parity_block_reasons()` does not block on unknown alone; resolver remains authoritative.
### Scope / independence
- Diff vs master: `master_parity_gate.py`, `gitea_mcp_server.py`, `tests/conftest.py`, `tests/test_master_parity_gate.py` only — scoped to #610.
- Author handoff #14074 cites head `57a52b1`; matches live PR head.
- Base `master`; mergeable; author != reviewer.
## Independent validation ledger
Worktree `branches/review-pr-788` @ `57a52b1`:
| Suite | Result |
|-------|--------|
| `tests/test_master_parity_gate.py` | **40 passed** |
| `test_mcp_server` + `test_runtime_clarity` + `test_resolve_task_capability` | **244 passed, 1 failed** |
| Full suite | **4262 passed, 11 failed, 6 skipped, 493 subtests** |
**Baseline classification:** the single focused failure (`TestPreflightVerification::test_declared_clean_task_worktree_ignores_control_checkout_violation`) and all 11 full-suite failures reproduce on master `620ed6e9a955` (detached baseline worktree). **Baseline, not branch-caused.** Prior review head had 115 failures; current head restores parity with master failure set.
Manual hermetic/production path probes: PRODUCTION_PROBE_OK, HERMETIC_OK, CLEAR_TRUE_OK, PYTEST_ENV_OK, OVERRIDE_OK.
## Canonical PR State
STATE: approved
WHO_IS_NEXT: merger
NEXT_ACTION: merge PR #788 at exact head 57a52b1a9933a7093498d988911f90a325615714 through sanctioned merger lease and merge path; do not self-merge from this reviewer session
NEXT_PROMPT:
```text
Merge PR #788 (Closes #610) on prgs Scaled-Tech-Consulting/Gitea-Tools as prgs-merger.
Pin head 57a52b1a9933a7093498d988911f90a325615714; confirm independent APPROVE from sysadmin at that head; revalidate checks and mergeability; merge via sanctioned MCP only.
```
WHAT_HAPPENED: Independent re-review of remediating head 57a52b1 found F1/F2 fully resolved; production live-remote probe intact; suite hermetic isolation correct; remaining failures are master baseline.
WHY: Author remediated non-hermetic ls-remote that broke feature-worktree test runs.
ISSUE: #610
HEAD_SHA: 57a52b1a9933a7093498d988911f90a325615714
REVIEW_STATUS: approved
MERGE_READY: true
BLOCKERS: none
VALIDATION: Independent focused and full suite as above; production vs hermetic path proofs; review 496 findings closed
NATIVE_REVIEW_PROOF: transport=native_mcp; entrypoint=mcp_server; token_fingerprint=73c82cf6a9dcf507
LAST_UPDATED_BY: sysadmin / prgs-reviewer
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
sysadmin
Closes #610
Head
324b4b3e9312840b44d1587d8b9142a76ca5ef74Publication recovery only: existing commits on
feat/issue-610-live-remote-paritywere already present onprgsand matched local HEAD. No amend, squash, rebase, or force was performed. This PR opens the missing review surface.Problem
gitea_assess_master_paritycompared only the daemon startup commit against the local on-disk HEAD. When the checkout was not updated, parity reported green even though live remote master had advanced, so a stale daemon could claim a mutation-safe result while running outdated capability gates (observed during PR #592 recovery, where the resolver correctly required restart but parity saidin_parity=true).Implementation summary
master_parity_gate.assess_master_parity()accepts optionallive_remote_headand reports three commits distinctly (daemon_start_head,local_head,live_remote_head) pluslive_known/live_stale/mutation_safe.parity_block_reasons()blocks mutations on live-staleness; read-only diagnostics stay unblocked when remote is unknown.parity_resolver_disagreement()provides a typed fail-closed blocker naming the capability resolver as authoritative when it requires restart while parity looks locally green.read_remote_master_head()best-effort live remote tip with 60s TTL; tests can pin viaGITEA_TEST_LIVE_REMOTE_HEAD.gitea_assess_master_parity/ runtime context.Files
master_parity_gate.pygitea_mcp_server.pytests/test_master_parity_gate.pyTest evidence (from authoring commit message; not re-run in this publisher)
test_configAuthIntegration /test_credentialsGetCredentials were baseline-proven keychain/env failures identical on the unmodified base (not branch-caused).Known integration note for reviewer
Branch head is from 2026-07-09 and is behind current master. Merge may need author
update_branch_by_mergeafter review; this publisher does not rewrite history or resolve conflicts.Publisher execution identity
grok-pr-publisher-prod-v1grok-pr-publisher-prod-v1-20260721T232804Z-pid18507-7b750edafeat/issue-610-live-remote-parity@324b4b3e9312840b44d1587d8b9142a76ca5ef74CTH: Author Handoff
Status: published_awaiting_independent_review
Next owner: reviewer (prgs-reviewer / sysadmin)
Current blocker: none for publication; branch is behind master and may need author update_branch_by_merge after review if conflicts appear
Decision: PR #788 created for completed committed work on issue #610; no code rewritten
Proof: gitea_create_pr → #788; head_sha
324b4b3e93matches local and prgs remote branch; base master; Closes #610; exactly one open PRNext action: independent formal review at exact head
324b4b3e93Ready-to-paste prompt: see NEXT_PROMPT below
Canonical Issue State
STATE: pr_open_awaiting_review
WHO_IS_NEXT: reviewer
NEXT_ACTION: formal independent review of PR #788 at head
324b4b3e93NEXT_PROMPT:
WHAT_HAPPENED: Publication recovery run adopted existing remote branch feat/issue-610-live-remote-parity @
324b4b3e93(already published; no push), locked issue #610 via own-branch adoption, and created PR #788 targeting master with Closes #610. Applied status:pr-open. No implementation, amend, rebase, or force-push.WHY: Completed committed work for open issue #610 was ahead of prgs/master with no PR; publisher recovers missing review surface only.
RELATED_PRS: #788 (this PR); #592 (historical incident context only, already merged)
BLOCKERS: none for review handoff; branch is 192 commits behind master so merge may later need author update_branch_by_merge
VALIDATION: Authoring-commit evidence: 13 new tests; full suite 2423 passed; 8 baseline keychain/env failures identical on unmodified base (not branch-caused). Publisher did not re-run the suite.
LAST_UPDATED_BY: jcwalker3 (prgs-author)
LLM_LOCK_ID: grok-pr-publisher-prod-v1
LLM_EXECUTION_ID: grok-pr-publisher-prod-v1-20260721T232804Z-pid18507-7b750eda
REPOSITORY: Scaled-Tech-Consulting/Gitea-Tools
ISSUE: #610
PR: #788
HEAD_SHA:
324b4b3e93BASE_BRANCH: master
BASE_OR_MERGE_SHA:
35e94e107cACTING_ROLE: author
ACTING_IDENTITY: jcwalker3
repo: Scaled-Tech-Consulting/Gitea-Tools
pr: #788
issue: #610
reviewer_identity: sysadmin
profile: prgs-reviewer
session_id: 21098-3d496ddf679b
worktree: /Users/jasonwalker/Development/Gitea-Tools/branches/review-pr-788
phase: claimed
candidate_head:
324b4b3e93target_branch: master
target_branch_sha:
35e94e107clast_activity: 2026-07-22T00:30:18Z
expires_at: 2026-07-22T00:40:18Z
blocker: none
repo: Scaled-Tech-Consulting/Gitea-Tools
pr: #788
issue: #610
reviewer_identity: sysadmin
profile: prgs-reviewer
session_id: 21098-f5dd578979ec
worktree: /Users/jasonwalker/Development/Gitea-Tools/branches/review-pr-788
phase: claimed
candidate_head:
324b4b3e93target_branch: master
target_branch_sha:
35e94e107clast_activity: 2026-07-22T00:42:18Z
expires_at: 2026-07-22T00:52:18Z
blocker: none
Verdict: REQUEST_CHANGES — PR #788 @
324b4b3e9312840b44d1587d8b9142a76ca5ef74Reviewed independently by
sysadmin/prgs-reviewer(authorjcwalker3; no self-review). Lease21098-f5dd578979ec, worktreebranches/review-pr-788, targetmaster@35e94e1.The #610 design (distinguish daemon-start / local / live-remote heads; fail closed on live-stale; resolver authoritative) is sound and well-scoped, and the 37 new
test_master_parity_gate.pycases pass. One blocking defect:F1 (blocking) — live-remote parity read is not hermetic; 107 pre-existing tests broken
read_remote_master_head()performs a realgit ls-remotewheneverGITEA_TEST_LIVE_REMOTE_HEADis unset, and nothing pins that variable suite-wide (notests/conftest.pyentry; grep shows the only references are the new module/tests). Whenever the checkout HEAD differs from live remote master — i.e. every canonicalbranches/feature worktree, the exact environment this workflow requires all test runs to use —gitea_get_runtime_context/gitea_assess_master_parityreportlive_staleand flipsafe_next_action, breaking legacy suites that assert runtime-context content.Proof (venv python,
pytest -q):324b4b3full suite: 115 failed, 2316 passed.5201379(detached baseline worktree) full suite, identical command: 8 failed, 2404 passed — the 8 are the declared keychain/env baseline set (test_configAuthIntegration /test_credentialsGetCredentials).test_mcp_server.py test_runtime_clarity.py test_resolve_task_capability.py): branch 54 failed / 190 passed vs baseline 244 passed / 0 failed.test_get_runtime_context_reviewer: got"Daemon is stale relative to LIVE remote master (started 324b4b3e9312, live master 35e94e107c32); …"where the test asserts"None; ready for operations.".GITEA_TEST_LIVE_REMOTE_HEADoutside the tests does not rescue them: the failing tests usepatch.dict(os.environ, …, clear=True), which wipes the pin, so the code falls through to a real networkls-remoteinside unit tests.Required remediation: make the live-remote read hermetic for the test suite — e.g. a
conftest.pypin ofGITEA_TEST_LIVE_REMOTE_HEAD, or disabling the live probe under pytest unless explicitly opted in — and update the affected legacy expectations for the new runtime-context fields/safe_next_actionso the full suite is green (minus the 8 baseline failures) when run from abranches/worktree whose HEAD differs from live master.F2 (non-blocking) — unit tests gain a network dependency
With F1 unfixed, every cold-cache test run performs
git ls-remote(5s timeout) against the real remote; offline runs silently change behavior (live_known=False→mutation_safe=Falseyet unblocked). The hermetic fix for F1 should remove this too.Note (non-blocking) — semantics of unknown live state
mutation_safe=Falsewhen the live head is unknown, butparity_block_reasons()does not block on unknown — documented and defensible (resolver stays authoritative); calling it out so the choice is deliberate.Integration note
The branch is ~12 days behind master (
mergeable: truetoday). After F1/F2 remediation, expect an authorupdate_branch_by_mergesync before merge; the reviewer handoff in the PR body already anticipates this.Validation ledger
master_parity_gate.py(+188/−26 net),gitea_mcp_server.py(+73),tests/test_master_parity_gate.py(+198). Scoped, no secrets, no gate weakening found.tests/test_master_parity_gate.py: 37 passed.gitea_load_review_workflowhash263d0a6cb8a6, boundary clean; head pinned324b4b3; verdict recorded viagitea_mark_final_review_decision.Canonical PR State
STATE: changes-requested
WHO_IS_NEXT: author
NEXT_ACTION: Remediate F1 (hermetic live-remote read / conftest pin + updated legacy expectations) and F2 on the PR branch, re-run the full suite from a branches/ worktree, then request re-review.
NEXT_PROMPT:
WHAT_HAPPENED: Independent review of PR #788 @
324b4b3found the #610 live-remote parity design sound but non-hermetic: 107 pre-existing tests break whenever the checkout HEAD differs from live remote master (115 failed vs 8 baseline-identical failures at merge-base5201379).WHY: The live-remote
git ls-remoteprobe is unpinned in the test suite; canonical branches/ worktree test runs always differ from live master, so runtime-context/parity outputs flip to live_stale and break legacy assertions.ISSUE: #610
HEAD_SHA:
324b4b3e93REVIEW_STATUS: changes-requested
MERGE_READY: no
BLOCKERS: F1 (blocking) non-hermetic live-remote parity read breaking 107 tests; F2 (non-blocking) unit-test network dependency; branch ~12 days behind master (author update_branch_by_merge after fixes).
VALIDATION: Focused test_master_parity_gate.py 37 passed; focused 3-suite compare branch 54 failed/190 passed vs baseline 244 passed; full suite branch 115 failed/2316 passed vs pristine merge-base 8 failed/2404 passed (8 = declared keychain/env baseline set); diff inspected, scoped, no secrets.
LAST_UPDATED_BY: sysadmin / prgs-reviewer (independent reviewer; lease 21098-f5dd578979ec)
CTH: Author Handoff
Status: published_awaiting_independent_review
Next owner: reviewer (prgs-reviewer / sysadmin)
Current blocker: none for re-review; prior REQUEST_CHANGES (review 496 @
324b4b3) remediatedDecision: F1/F2 hermetic live-remote fix published; ready for independent re-review
Proof: gitea_commit_files →
57a52b1a9933a7093498d988911f90a325615714onfeat/issue-610-live-remote-parity; PR #788 head matches; branch already mergeable vs current master (prior update-by-merge at344dc41)Next action: independent formal re-review at exact head
57a52b1a9933a7093498d988911f90a325615714Ready-to-paste prompt: see NEXT_PROMPT below
Remediation of review 496 (F1/F2)
F1 (blocking) — hermetic live-remote reads
master_parity_gate.set_hermetic_test_mode(True)enabled suite-wide via autouse fixture intests/conftest.py.GITEA_TEST_LIVE_REMOTE_HEADis unset,read_remote_master_headreturnsNonewithoutgit ls-remote(module flag survivespatch.dict(..., clear=True)).PYTEST_CURRENT_TESTis set unlessGITEA_TEST_ALLOW_LIVE_REMOTE_PROBE.F2 (non-blocking) — no unit-test network dependency
ls-remotefrom the suite.Files:
master_parity_gate.py,tests/conftest.py,tests/test_master_parity_gate.py(+3 hermetic cases).Validation (from allocated worktree)
Run with
/Users/jasonwalker/Development/Gitea-Tools/venv/bin/python -m pytestat worktreebranches/author-issue-610-publish-lock@57a52b1:tests/test_master_parity_gate.py— 40 passedtest_mcp_server.py+test_runtime_clarity.py+test_resolve_task_capability.py) — 244 passed, 1 failedBaseline comparison: master
620ed6e9a955(worktreebranches/baseline-master-issue-790) fails the same single focused caseTestPreflightVerification::test_declared_clean_task_worktree_ignores_control_checkout_violation. The eleven full-suite failures match the baseline set documented on PR #794 (commit_payloads ×6, issue_702 F1/F2 ×2, that preflight case, post_merge_moot_lease, reconciler_supersession_close). Prior review head324b4b3had 115 failed; after F1 hermetic + master merge: 11 failed (baseline only). No new branch-caused failures.Canonical Issue State
STATE: pr_open_awaiting_review
WHO_IS_NEXT: reviewer
NEXT_ACTION: formal independent re-review of PR #788 at head
57a52b1a99NEXT_PROMPT:
WHAT_HAPPENED: Author remediates review 496 F1/F2 on issue #610 / PR #788. Published commit
57a52b1via gitea_commit_files from worktree branches/author-issue-610-publish-lock after renewing the expired exact-owner author lock. Prior head344dc41already included master merge; this commit adds hermetic test isolation only.WHY: REQUEST_CHANGES required hermetic live-remote parity reads so feature worktrees do not break 100+ legacy runtime-context tests.
RELATED_PRS: #788 (this PR)
BLOCKERS: none for re-review
VALIDATION: test_master_parity_gate 40 passed; focused 244 passed / 1 baseline fail; full 4262 passed / 11 baseline fail / 6 skipped; head read-back
57a52b1matches gitea_get_pr_review_feedback current_head_shaLAST_UPDATED_BY: jcwalker3 (prgs-author)
LLM_LOCK_ID: grok-pr-queue-author-788
LLM_EXECUTION_ID: grok-pr-queue-author-788-20260722T1036Z
REPOSITORY: Scaled-Tech-Consulting/Gitea-Tools
ISSUE: #610
PR: #788
HEAD_SHA:
57a52b1a99BASE_BRANCH: master
BASE_OR_MERGE_SHA:
620ed6e9a9ACTING_ROLE: author
ACTING_IDENTITY: jcwalker3
repo: Scaled-Tech-Consulting/Gitea-Tools
pr: #788
issue: #610
reviewer_identity: sysadmin
profile: prgs-reviewer
session_id: 89687-2a94cee47d5c
worktree: /Users/jasonwalker/Development/Gitea-Tools/branches/review-pr-788
phase: claimed
candidate_head:
57a52b1a99target_branch: master
target_branch_sha:
620ed6e9a9last_activity: 2026-07-22T10:43:50Z
expires_at: 2026-07-22T10:53:50Z
blocker: none
Verdict: APPROVE — PR #788 @
57a52b1a9933a7093498d988911f90a325615714Reviewed independently by
sysadmin/prgs-reviewer(authorjcwalker3; no self-review). Lease session89687-2a94cee47d5c/ comment 14084, worktreebranches/review-pr-788, targetmaster@620ed6e9a955.Re-evaluation of review 496
F1 (was blocking) — RESOLVED
Hermetic live-remote isolation is now suite-wide:
_HERMETIC_TEST_MODEdefault False (production probes).tests/conftest.pyautouse_hermetic_live_remote_master_headenables hermetic for pytest and disables infinally.PYTEST_CURRENT_TESTset) and no override /GITEA_TEST_ALLOW_LIVE_REMOTE_PROBE,read_remote_master_headreturnsNonewithoutgit ls-remote.patch.dict(..., clear=True)— independent proof: subprocess never called under cleared env.GITEA_TEST_ALLOW_LIVE_REMOTE_PROBEwhile still mocking subprocess.Independent production-path proof (no hermetic, no pytest env):
read_remote_master_headstill issuesgit ls-remoteonce. Server_current_master_parity()still wires live head intoassess_master_parity.F2 (was non-blocking) — RESOLVED
Unit tests no longer cold-cache network
ls-remoteunder suite hermetic mode.Non-blocking residual note (not a merge blocker)
Defense-in-depth uses
PYTEST_CURRENT_TEST. A production process whose environment was polluted with that variable would skip the live probe. That is not the normal production path (_HERMETIC_TEST_MODEdefaults false; MCP daemon is not under pytest). Documented for operators only.Unknown live-state semantics
Unchanged and still defensible:
mutation_safe=Falsewhen live unknown;parity_block_reasons()does not block on unknown alone; resolver remains authoritative.Scope / independence
master_parity_gate.py,gitea_mcp_server.py,tests/conftest.py,tests/test_master_parity_gate.pyonly — scoped to #610.57a52b1; matches live PR head.master; mergeable; author != reviewer.Independent validation ledger
Worktree
branches/review-pr-788@57a52b1:tests/test_master_parity_gate.pytest_mcp_server+test_runtime_clarity+test_resolve_task_capabilityBaseline classification: the single focused failure (
TestPreflightVerification::test_declared_clean_task_worktree_ignores_control_checkout_violation) and all 11 full-suite failures reproduce on master620ed6e9a955(detached baseline worktree). Baseline, not branch-caused. Prior review head had 115 failures; current head restores parity with master failure set.Manual hermetic/production path probes: PRODUCTION_PROBE_OK, HERMETIC_OK, CLEAR_TRUE_OK, PYTEST_ENV_OK, OVERRIDE_OK.
Canonical PR State
STATE: approved
WHO_IS_NEXT: merger
NEXT_ACTION: merge PR #788 at exact head
57a52b1a99through sanctioned merger lease and merge path; do not self-merge from this reviewer sessionNEXT_PROMPT:
WHAT_HAPPENED: Independent re-review of remediating head
57a52b1found F1/F2 fully resolved; production live-remote probe intact; suite hermetic isolation correct; remaining failures are master baseline.WHY: Author remediated non-hermetic ls-remote that broke feature-worktree test runs.
ISSUE: #610
HEAD_SHA:
57a52b1a99REVIEW_STATUS: approved
MERGE_READY: true
BLOCKERS: none
VALIDATION: Independent focused and full suite as above; production vs hermetic path proofs; review 496 findings closed
NATIVE_REVIEW_PROOF: transport=native_mcp; entrypoint=mcp_server; token_fingerprint=73c82cf6a9dcf507
LAST_UPDATED_BY: sysadmin / prgs-reviewer
adopted_at: 2026-07-22T10:51:07Z
adopted_by_identity: sysadmin
adopted_by_profile: prgs-merger
adopted_from_session_id: 89687-2a94cee47d5c
adopted_from_profile: prgs-reviewer
adopted_from_reviewer_identity: sysadmin
adopted_from_comment_id: 14084
adoption_reason: merger-handoff-approved-head
repo: Scaled-Tech-Consulting/Gitea-Tools
pr: #788
issue: #610
reviewer_identity: sysadmin
profile: prgs-merger
session_id: 18017-014f996adf50
worktree: /Users/jasonwalker/Development/Gitea-Tools
phase: adopted
candidate_head:
57a52b1a99target_branch: master
target_branch_sha:
620ed6e9a9last_activity: 2026-07-22T10:51:07Z
expires_at: 2026-07-22T11:01:07Z
blocker: none
Stale #332 review-decision lock cleanup (#594)
Status: APPLIED
sysadminprgs-merger2026-07-22T10:51:57.674278+00:00approveon PR feat: make master-parity live-remote aware so a stale daemon fails closed (Closes #610) (#788)closed(merged=True)5032965e3a39654f1d65b2a58f25931bc72ff17b1prgs-reviewerManual deletion of session-state files is not the workflow.
This path only clears a lock when the referenced PR is merged/closed.
repo: Scaled-Tech-Consulting/Gitea-Tools
pr: #788
issue: #610
reviewer_identity: sysadmin
profile: prgs-merger
session_id: 18017-014f996adf50
worktree: /Users/jasonwalker/Development/Gitea-Tools
phase: released
candidate_head:
57a52b1a99target_branch: master
target_branch_sha:
620ed6e9a9last_activity: 2026-07-22T10:52:11Z
expires_at: 2026-07-22T11:02:11Z
blocker: manual-release