Scaled-Tech-Consulting/Gitea-Tools
2
0
Fork 0
Code Issues 73 Pull Requests 9 Actions Packages Projects Releases Wiki Activity

Harden workflow: block unattributed root WIP and pytest-disabled production guards #683

New Issue
Closed
opened 2026-07-12 12:04:11 -05:00 by jcwalker3 · 7 comments
jcwalker3 commented 2026-07-12 12:04:11 -05:00
Owner
Copy Link
Copy Source

Context

Spawned from preservation issue #681 (dirty control-checkout WIP left by the PR #680 / issue #604 review session).

A reviewer session investigating anti-stomp / preflight behavior under pytest edited production modules on the root control checkout without an owning issue-backed branches/ worktree. The WIP was later preserved only (not endorsed) as:

  • branch: chore/issue-681-preserve-review-session-wip
  • commit: 300a4ca08fee3a195efa723d17eb6c64a3ef7a4f

300a4ca is evidence, not merge-ready code. Do not cherry-pick it wholesale. Do not merge the preservation branch.

Root cause (process)

  1. Mutation of source/tests occurred on the control/root checkout instead of an issue-bound worktree under branches/.
  2. No durable owning issue was selected before out-of-scope diagnostic edits.
  3. Test-isolation attempts weakened production guards (test-mode early-return; porcelain filtering of *.py; broad patches of root/branches guards) rather than proving production entrypoints remain fail-closed under pytest.
  4. Workflow/tool failures during review (EOF/#584, config drift/#672) did not block further root source edits before durable recording.

Related but not owners of this work:

  • #681 — preservation/recovery of the unattributed WIP (control clean + durable capture). Does not authorize adopting the rejected hunks.
  • #604 / PR #680 — anti-stomp preflight product work; must not absorb this WIP.
  • #675 / PR #676 — prior residual preflight/test-isolation remediations (already closed/merged); different patterns.
  • #584, #672 — transport EOF / Antigravity config drift during the same incident cluster.
  • #618 — missing author worktree binding consistency (sibling guardrail, different defect).

Controller disposition of preserved hunks (300a4ca)

File Disposition
gitea_mcp_server.py — early-return in _verify_role_mutation_workspace when _preflight_in_test_mode() REJECT — pytest-only bypass of production workspace binding / anti-stomp path
issue_lock_worktree.py — strip porcelain lines ending in .py under pytest/unittest REJECT — conceals real dirty tracked Python files from safety gates
tests/test_preflight_read_survival.py — patch _enforce_branches_only_author_mutation + _enforce_root_checkout_guard Retain concept only if rewritten — unit isolation allowed only when at least one real-entrypoint integration test still proves production guards remain active

No separate product issue is opened to reimplement the rejected production bypasses. Valid implementable work is this hardening issue (process + regression proof).

Acceptance criteria (required)

  1. Root mutation block — Source or test edits in the root/control checkout are blocked, including temporary diagnostic edits.
  2. Scope ownership first — An edit outside the active issue’s scope requires selecting or creating an owning issue first.
  3. Issue-backed worktree bind — The session must bind to an issue-backed worktree under branches/ before mutation.
  4. Pytest cannot disable production guards — Pytest cannot disable, bypass, or conceal production root-dirtiness and anti-stomp guards (no production early-return that skips _verify_role_mutation_workspace / root / branches enforcement solely because tests are running).
  5. Porcelain integrity — Git porcelain results used by safety gates cannot silently omit tracked dirty source files (no *.py filtering in read_worktree_git_state or equivalents).
  6. Explicit test isolation + entrypoint proof — Test isolation must be explicit (fixtures/env/mocks scoped to the test) and must preserve at least one real-entrypoint integration test proving the production guard remains active under pytest (e.g. force-on patterns like GITEA_TEST_FORCE_ANTI_STOMP=1 where applicable).
  7. Typed blocker — The blocker must return a typed reason and an exact next action.
  8. Durable failure recording — Workflow/tool failures that trigger new source edits must be recorded durably (issue comment / ledger) before work continues.

Required regression tests

  • Assert production _verify_role_mutation_workspace (or successor) does not short-circuit solely via _preflight_in_test_mode() in a way that skips root/workspace binding when forced-on flags request production behavior.
  • Assert read_worktree_git_state / porcelain consumers still surface dirty *.py files under pytest.
  • Assert at least one mutation entrypoint (e.g. comment_issue or create_issue) fails closed on dirty root / missing branches worktree with blocker_kind + exact_next_action while pytest is running.
  • Negative tests that would have accepted the rejected 300a4ca patterns must fail (or not exist as product code).
  • Optional: session/workflow docs or ledger example for “root WIP unattributed → stop → preserve or claim issue”.

Explicit non-goals

  • Do not merge chore/issue-681-preserve-review-session-wip or any alleged scope-hardening branch as a product fix.
  • Do not cherry-pick 300a4ca.
  • Do not absorb changes into #604 / PR #680.
  • Do not “fix” suite determinism by masking dirty-root state in production modules.

Implementation direction

  • Prefer fail-closed preflight / root-checkout / branches-only enforcement with typed blockers.
  • Prefer test isolation via env fixtures (GITEA_TEST_PORCELAIN, forced-on flags) and test-local mocks — never by filtering porcelain in production readers.
  • When diagnostic source edits are needed, bind branches/issue-NNN-* first and claim the issue.

Handoff

WHO_IS_NEXT: author
STATE: ready-for-author

See issue comment for the ready-to-run AUTHOR prompt.

## Context Spawned from preservation issue **#681** (dirty control-checkout WIP left by the PR **#680** / issue **#604** review session). A reviewer session investigating anti-stomp / preflight behavior under pytest edited **production modules on the root control checkout** without an owning issue-backed `branches/` worktree. The WIP was later preserved only (not endorsed) as: * branch: `chore/issue-681-preserve-review-session-wip` * commit: `300a4ca08fee3a195efa723d17eb6c64a3ef7a4f` **`300a4ca` is evidence, not merge-ready code. Do not cherry-pick it wholesale. Do not merge the preservation branch.** ## Root cause (process) 1. Mutation of source/tests occurred on the control/root checkout instead of an issue-bound worktree under `branches/`. 2. No durable owning issue was selected before out-of-scope diagnostic edits. 3. Test-isolation attempts weakened production guards (test-mode early-return; porcelain filtering of `*.py`; broad patches of root/branches guards) rather than proving production entrypoints remain fail-closed under pytest. 4. Workflow/tool failures during review (EOF/#584, config drift/#672) did not block further root source edits before durable recording. Related but **not** owners of this work: * **#681** — preservation/recovery of the unattributed WIP (control clean + durable capture). Does **not** authorize adopting the rejected hunks. * **#604** / **PR #680** — anti-stomp preflight product work; must not absorb this WIP. * **#675** / **PR #676** — prior residual preflight/test-isolation remediations (already closed/merged); different patterns. * **#584**, **#672** — transport EOF / Antigravity config drift during the same incident cluster. * **#618** — missing author worktree binding consistency (sibling guardrail, different defect). ## Controller disposition of preserved hunks (`300a4ca`) | File | Disposition | |------|-------------| | `gitea_mcp_server.py` — early-return in `_verify_role_mutation_workspace` when `_preflight_in_test_mode()` | **REJECT** — pytest-only bypass of production workspace binding / anti-stomp path | | `issue_lock_worktree.py` — strip porcelain lines ending in `.py` under pytest/unittest | **REJECT** — conceals real dirty tracked Python files from safety gates | | `tests/test_preflight_read_survival.py` — patch `_enforce_branches_only_author_mutation` + `_enforce_root_checkout_guard` | **Retain concept only if rewritten** — unit isolation allowed only when at least one real-entrypoint integration test still proves production guards remain active | No separate product issue is opened to reimplement the rejected production bypasses. Valid implementable work is **this hardening issue** (process + regression proof). ## Acceptance criteria (required) 1. **Root mutation block** — Source or test edits in the root/control checkout are blocked, including temporary diagnostic edits. 2. **Scope ownership first** — An edit outside the active issue’s scope requires selecting or creating an owning issue first. 3. **Issue-backed worktree bind** — The session must bind to an issue-backed worktree under `branches/` before mutation. 4. **Pytest cannot disable production guards** — Pytest cannot disable, bypass, or conceal production root-dirtiness and anti-stomp guards (no production early-return that skips `_verify_role_mutation_workspace` / root / branches enforcement solely because tests are running). 5. **Porcelain integrity** — Git porcelain results used by safety gates cannot silently omit tracked dirty source files (no `*.py` filtering in `read_worktree_git_state` or equivalents). 6. **Explicit test isolation + entrypoint proof** — Test isolation must be explicit (fixtures/env/mocks scoped to the test) and must preserve at least one real-entrypoint integration test proving the production guard remains active under pytest (e.g. force-on patterns like `GITEA_TEST_FORCE_ANTI_STOMP=1` where applicable). 7. **Typed blocker** — The blocker must return a typed reason and an exact next action. 8. **Durable failure recording** — Workflow/tool failures that trigger new source edits must be recorded durably (issue comment / ledger) before work continues. ## Required regression tests * Assert production `_verify_role_mutation_workspace` (or successor) does **not** short-circuit solely via `_preflight_in_test_mode()` in a way that skips root/workspace binding when forced-on flags request production behavior. * Assert `read_worktree_git_state` / porcelain consumers still surface dirty `*.py` files under pytest. * Assert at least one mutation entrypoint (e.g. `comment_issue` or `create_issue`) fails closed on dirty root / missing branches worktree with `blocker_kind` + `exact_next_action` while pytest is running. * Negative tests that would have accepted the rejected `300a4ca` patterns must fail (or not exist as product code). * Optional: session/workflow docs or ledger example for “root WIP unattributed → stop → preserve or claim issue”. ## Explicit non-goals * Do not merge `chore/issue-681-preserve-review-session-wip` or any alleged scope-hardening branch as a product fix. * Do not cherry-pick `300a4ca`. * Do not absorb changes into #604 / PR #680. * Do not “fix” suite determinism by masking dirty-root state in production modules. ## Implementation direction * Prefer fail-closed preflight / root-checkout / branches-only enforcement with typed blockers. * Prefer test isolation via env fixtures (`GITEA_TEST_PORCELAIN`, forced-on flags) and **test-local** mocks — never by filtering porcelain in production readers. * When diagnostic source edits are needed, bind `branches/issue-NNN-*` first and claim the issue. ## Handoff **WHO_IS_NEXT:** author **STATE:** ready-for-author See issue comment for the ready-to-run AUTHOR prompt.
jcwalker3 commented 2026-07-12 12:04:56 -05:00
Author
Owner
Copy Link
Copy Source

Canonical Issue State

STATE:
ready-for-author

WHO_IS_NEXT:
author

NEXT_ACTION:
Bind issue-backed worktree under branches/; implement ACs from clean master; open linked PR Closes #683; stop for independent review

NEXT_PROMPT:

Role: AUTHOR
Repo: prgs / Scaled-Tech-Consulting / Gitea-Tools
Issue: #683 (Closes #683)
Evidence only: #681 branch chore/issue-681-preserve-review-session-wip @ 300a4ca — do NOT cherry-pick wholesale; do NOT merge preserve branch

1. Identity: gitea_whoami + gitea_resolve_task_capability on prgs-author; confirm allowed_in_current_session.
2. Bind clean issue-backed worktree under branches/ (e.g. branches/issue-683-*) from current master; set worktree_path / GITEA_AUTHOR_WORKTREE. Never mutate the control root checkout.
3. Reimplement ONLY the hardening ACs in #683 from clean canonical sources. REJECT production early-return in _verify_role_mutation_workspace under test mode. REJECT porcelain filtering of *.py in issue_lock_worktree.read_worktree_git_state.
4. If test_preflight_read_survival needs isolation, use explicit test-local fixtures/env; keep >=1 real-entrypoint integration test proving root/branches/anti-stomp production guards remain active under pytest.
5. Add behavioral regression tests listed on #683 body.
6. Open a linked PR with Closes #683; stop for independent REVIEWER. Do not self-review or merge. Do not assign reconciler until independent approval exists.
7. Optionally note on #681 that product disposition remains reject for 300a4ca production hunks and that #683 owns hardening.

WHAT_HAPPENED:
CONTROLLER created this issue as durable owner of process + regression hardening after classifying #681 preserved commit 300a4ca. Rejected production bypass hunks; no second issue opened to reimplement those bypasses. Posted classification on #681.

WHY:
Unattributed root WIP and pytest-disabled production guards are a workflow-hardening failure distinct from #681 preservation and from #604/#680 product anti-stomp. Without durable ownership, rejected evidence can be smuggled into unrelated PRs.

RELATED_PRS:
#680 (incident parent review; do not absorb WIP); #676 (prior residual isolation; closed; different patterns)

BLOCKERS:
none for author start; must not use 300a4ca as implementation base

VALIDATION:

  • Issue #683 created with type:guardrail status:ready and 8 ACs in body
  • #681 comment posted with full hunk classification and reject dispositions
  • Control checkout clean; preserve branch left unmerged
  • prgs-author jcwalker3 create_issue + comment_issue allowed this session

LAST_UPDATED_BY:
jcwalker3 / CONTROLLER / 2026-07-12

Root cause (summary)

Root control-checkout source/test edits during PR #680 review without issue-backed branches/ bind; diagnostic isolation weakened production guards instead of proving entrypoints fail closed under pytest.

Required regression tests (reminder)

  1. No production short-circuit of workspace verify solely via _preflight_in_test_mode() when force-on production behavior is required.
  2. Porcelain still surfaces dirty *.py under pytest.
  3. Mutation entrypoint fails closed on dirty root / missing branches worktree with typed blocker_kind + exact_next_action under pytest.
  4. Rejected 300a4ca production patterns must not land.
## Canonical Issue State STATE: ready-for-author WHO_IS_NEXT: author NEXT_ACTION: Bind issue-backed worktree under branches/; implement ACs from clean master; open linked PR Closes #683; stop for independent review NEXT_PROMPT: ```text Role: AUTHOR Repo: prgs / Scaled-Tech-Consulting / Gitea-Tools Issue: #683 (Closes #683) Evidence only: #681 branch chore/issue-681-preserve-review-session-wip @ 300a4ca — do NOT cherry-pick wholesale; do NOT merge preserve branch 1. Identity: gitea_whoami + gitea_resolve_task_capability on prgs-author; confirm allowed_in_current_session. 2. Bind clean issue-backed worktree under branches/ (e.g. branches/issue-683-*) from current master; set worktree_path / GITEA_AUTHOR_WORKTREE. Never mutate the control root checkout. 3. Reimplement ONLY the hardening ACs in #683 from clean canonical sources. REJECT production early-return in _verify_role_mutation_workspace under test mode. REJECT porcelain filtering of *.py in issue_lock_worktree.read_worktree_git_state. 4. If test_preflight_read_survival needs isolation, use explicit test-local fixtures/env; keep >=1 real-entrypoint integration test proving root/branches/anti-stomp production guards remain active under pytest. 5. Add behavioral regression tests listed on #683 body. 6. Open a linked PR with Closes #683; stop for independent REVIEWER. Do not self-review or merge. Do not assign reconciler until independent approval exists. 7. Optionally note on #681 that product disposition remains reject for 300a4ca production hunks and that #683 owns hardening. ``` WHAT_HAPPENED: CONTROLLER created this issue as durable owner of process + regression hardening after classifying #681 preserved commit 300a4ca. Rejected production bypass hunks; no second issue opened to reimplement those bypasses. Posted classification on #681. WHY: Unattributed root WIP and pytest-disabled production guards are a workflow-hardening failure distinct from #681 preservation and from #604/#680 product anti-stomp. Without durable ownership, rejected evidence can be smuggled into unrelated PRs. RELATED_PRS: #680 (incident parent review; do not absorb WIP); #676 (prior residual isolation; closed; different patterns) BLOCKERS: none for author start; must not use 300a4ca as implementation base VALIDATION: - Issue #683 created with type:guardrail status:ready and 8 ACs in body - #681 comment posted with full hunk classification and reject dispositions - Control checkout clean; preserve branch left unmerged - prgs-author jcwalker3 create_issue + comment_issue allowed this session LAST_UPDATED_BY: jcwalker3 / CONTROLLER / 2026-07-12 ### Root cause (summary) Root control-checkout source/test edits during PR #680 review without issue-backed branches/ bind; diagnostic isolation weakened production guards instead of proving entrypoints fail closed under pytest. ### Required regression tests (reminder) 1. No production short-circuit of workspace verify solely via `_preflight_in_test_mode()` when force-on production behavior is required. 2. Porcelain still surfaces dirty `*.py` under pytest. 3. Mutation entrypoint fails closed on dirty root / missing branches worktree with typed `blocker_kind` + `exact_next_action` under pytest. 4. Rejected 300a4ca production patterns must not land.
jcwalker3 added status:in-progress and removed status:ready labels 2026-07-12 12:09:12 -05:00
jcwalker3 commented 2026-07-12 12:09:12 -05:00
Author
Owner
Copy Link
Copy Source

Issue claim heartbeat

<!-- gitea-issue-claim-heartbeat:v1 --> **Issue claim heartbeat** - kind: claim - issue: #683 - branch: fix/issue-683-workflow-guard-hardening - phase: claimed - profile: prgs-author - pr: none - blocker: none - next_action: create worktree and begin implementation
jcwalker3 added status:pr-open and removed status:in-progress labels 2026-07-12 12:16:26 -05:00
jcwalker3 commented 2026-07-12 12:17:00 -05:00
Author
Owner
Copy Link
Copy Source

Canonical Issue State

STATE:
author-complete — PR open; waiting independent review

WHO_IS_NEXT:
reviewer

NEXT_ACTION:
Independent REVIEWER validates PR #684 against #683 ACs; do not self-review or merge

NEXT_PROMPT:

Role: REVIEWER (must differ from PR author jcwalker3 / prgs-author)
Repo: prgs / Scaled-Tech-Consulting / Gitea-Tools
PR: #684  |  Issue: #683 (Closes #683)
Base: master
Head: fix/issue-683-workflow-guard-hardening @ d302602

1. gitea_whoami + gitea_resolve_task_capability on prgs-reviewer; confirm allowed_in_current_session.
2. Inventory live PR #684 (title/body/files/head SHA). Pin expected_head_sha.
3. Create review worktree under branches/ for PR #684; never mutate control root.
4. Validate scope is #683 only — reject any cherry-pick of 300a4ca, any absorb of #604/#680, any porcelain *.py filter, any test-mode early-return in _verify_role_mutation_workspace.
5. Confirm ACs: root mutation block; scope ownership; issue-backed worktree bind; pytest cannot disable production guards when force-on; porcelain integrity; typed blocker + exact_next_action; durable failure recording helper; entrypoint proof before side effects.
6. Run focused #683 tests + root/contamination/preflight/namespace suites + full suite from the review worktree.
7. Post independent verdict (APPROVE or REQUEST_CHANGES) with proof. Do not merge. Do not self-author fixes on this PR unless operator reassigns.

WHAT_HAPPENED:
AUTHOR claimed #683, bound branches/issue-683-workflow-guard-hardening from clean master, implemented process hardening only (no cherry-pick of 300a4ca; no absorb of #604/PR #680). Opened PR #684 with Closes #683.

Implementation summary:

  • New workflow_scope_guard.py: issue scope ownership, root diagnostic block, typed blocker_kind + exact_next_action, GITEA_TEST_FORCE_PRODUCTION_GUARDS force-on, durable failure ledger helpers
  • gitea_mcp_server.py: production root/branches/scope via verify_preflight_purity and _enforce_issue_scope_guard; no test-mode early-return in _verify_role_mutation_workspace; typed blocks on gitea_create_issue and gitea_create_issue_comment; actual-profile reconciler exemption (#540)
  • issue_lock_worktree.read_worktree_git_state unchanged (no *.py porcelain filter)

Exact guard locations/entrypoints:

  • workflow_scope_guard.py (assessors + typed blockers)
  • gitea_mcp_server.verify_preflight_purity (purity-order + force-on production path)
  • gitea_mcp_server._enforce_issue_scope_guard / _enforce_root_checkout_guard / _enforce_branches_only_author_mutation
  • gitea_mcp_server._verify_role_mutation_workspace (namespace bind; no pytest early-return)
  • gitea_create_issue / gitea_create_issue_comment (typed blockers before API side effects)
  • issue_lock_worktree.read_worktree_git_state (truthful porcelain)

Proof pytest cannot disable production enforcement:

  • Default unit isolation may skip purity-order only; GITEA_TEST_FORCE_PRODUCTION_GUARDS=1 forces root+branches+scope under pytest
  • Real entrypoint test: dirty root + force-on + gitea_create_issue_comment → success=False, blocker_kind set, API not called
  • Static: no _preflight_in_test_mode early-return in workspace verify; no porcelain py strip under pytest
  • Monkeypatching branches-only + root helpers still leaves composed dirty-root assessment blocking

WHY:
Unattributed root WIP and pytest-disabled production guards are a distinct workflow-hardening failure from #681 preservation and #604/#680 product anti-stomp. #683 owns durable process + regression proof.

RELATED_PRS:
#684 (this author PR; Closes #683); #680 (incident parent anti-stomp — do not absorb); #681 preserve branch evidence only — do not merge

BLOCKERS:
none for reviewer start

VALIDATION:

  • Focused #683 suite passed (tests/test_issue_683_workflow_scope_guards.py)
  • Root checkout, author mutation, namespace binding, preflight-read survival, workspace guards, stable contamination suites passed
  • Full suite: 2615 passed, 6 skipped
  • compileall OK on changed modules
  • Control checkout clean on master @ 22698c1 throughout; worktree branches/issue-683-workflow-guard-hardening @ d302602

LAST_UPDATED_BY:
jcwalker3 / prgs-author / 2026-07-12

## Canonical Issue State STATE: author-complete — PR open; waiting independent review WHO_IS_NEXT: reviewer NEXT_ACTION: Independent REVIEWER validates PR #684 against #683 ACs; do not self-review or merge NEXT_PROMPT: ```text Role: REVIEWER (must differ from PR author jcwalker3 / prgs-author) Repo: prgs / Scaled-Tech-Consulting / Gitea-Tools PR: #684 | Issue: #683 (Closes #683) Base: master Head: fix/issue-683-workflow-guard-hardening @ d302602 1. gitea_whoami + gitea_resolve_task_capability on prgs-reviewer; confirm allowed_in_current_session. 2. Inventory live PR #684 (title/body/files/head SHA). Pin expected_head_sha. 3. Create review worktree under branches/ for PR #684; never mutate control root. 4. Validate scope is #683 only — reject any cherry-pick of 300a4ca, any absorb of #604/#680, any porcelain *.py filter, any test-mode early-return in _verify_role_mutation_workspace. 5. Confirm ACs: root mutation block; scope ownership; issue-backed worktree bind; pytest cannot disable production guards when force-on; porcelain integrity; typed blocker + exact_next_action; durable failure recording helper; entrypoint proof before side effects. 6. Run focused #683 tests + root/contamination/preflight/namespace suites + full suite from the review worktree. 7. Post independent verdict (APPROVE or REQUEST_CHANGES) with proof. Do not merge. Do not self-author fixes on this PR unless operator reassigns. ``` WHAT_HAPPENED: AUTHOR claimed #683, bound branches/issue-683-workflow-guard-hardening from clean master, implemented process hardening only (no cherry-pick of 300a4ca; no absorb of #604/PR #680). Opened PR #684 with Closes #683. Implementation summary: - New workflow_scope_guard.py: issue scope ownership, root diagnostic block, typed blocker_kind + exact_next_action, GITEA_TEST_FORCE_PRODUCTION_GUARDS force-on, durable failure ledger helpers - gitea_mcp_server.py: production root/branches/scope via verify_preflight_purity and _enforce_issue_scope_guard; no test-mode early-return in _verify_role_mutation_workspace; typed blocks on gitea_create_issue and gitea_create_issue_comment; actual-profile reconciler exemption (#540) - issue_lock_worktree.read_worktree_git_state unchanged (no *.py porcelain filter) Exact guard locations/entrypoints: - workflow_scope_guard.py (assessors + typed blockers) - gitea_mcp_server.verify_preflight_purity (purity-order + force-on production path) - gitea_mcp_server._enforce_issue_scope_guard / _enforce_root_checkout_guard / _enforce_branches_only_author_mutation - gitea_mcp_server._verify_role_mutation_workspace (namespace bind; no pytest early-return) - gitea_create_issue / gitea_create_issue_comment (typed blockers before API side effects) - issue_lock_worktree.read_worktree_git_state (truthful porcelain) Proof pytest cannot disable production enforcement: - Default unit isolation may skip purity-order only; GITEA_TEST_FORCE_PRODUCTION_GUARDS=1 forces root+branches+scope under pytest - Real entrypoint test: dirty root + force-on + gitea_create_issue_comment → success=False, blocker_kind set, API not called - Static: no _preflight_in_test_mode early-return in workspace verify; no porcelain py strip under pytest - Monkeypatching branches-only + root helpers still leaves composed dirty-root assessment blocking WHY: Unattributed root WIP and pytest-disabled production guards are a distinct workflow-hardening failure from #681 preservation and #604/#680 product anti-stomp. #683 owns durable process + regression proof. RELATED_PRS: #684 (this author PR; Closes #683); #680 (incident parent anti-stomp — do not absorb); #681 preserve branch evidence only — do not merge BLOCKERS: none for reviewer start VALIDATION: - Focused #683 suite passed (tests/test_issue_683_workflow_scope_guards.py) - Root checkout, author mutation, namespace binding, preflight-read survival, workspace guards, stable contamination suites passed - Full suite: 2615 passed, 6 skipped - compileall OK on changed modules - Control checkout clean on master @ 22698c1 throughout; worktree branches/issue-683-workflow-guard-hardening @ d302602 LAST_UPDATED_BY: jcwalker3 / prgs-author / 2026-07-12
jcwalker3 commented 2026-07-12 12:17:36 -05:00
Author
Owner
Copy Link
Copy Source

Issue claim heartbeat

<!-- gitea-issue-claim-heartbeat:v1 --> **Issue claim heartbeat** - kind: progress - issue: #683 - branch: fix/issue-683-workflow-guard-hardening - phase: pr-open - profile: prgs-author - pr: #684 - blocker: none - next_action: independent reviewer validates PR #684
sysadmin commented 2026-07-12 18:55:30 -05:00
Owner
Copy Link
Copy Source

[THREAD STATE LEDGER] Issue #683 — PR #684 carries a formal APPROVE verdict at head d302602; merger workflow is next

What is true now:

  • Server-side decision state: PR #684 (Closes #683) carries a formal APPROVE verdict by sysadmin/prgs-reviewer recorded via the review API; read-back shows approval_visible=true, approval_at_current_head=true at head d302602567, has_blocking_change_requests=false
  • Local verdict/state: matches server; independent review evidence in PR #684 comment 10606 at the same head
  • Latest known validation: focused 19 passed; changed suites 30 passed; guard/lease/review/merge suites 155 passed; full suite 2615 passed / 6 skipped; compileall OK; four findings non-blocking

What changed:

  • Formal verdict recorded in a fresh client-managed reviewer run after operator reconnect (fresh lease comment 10670, session 76849-c083d5c9e7da; verdict read back at 2026-07-12T18:53:34-05:00; PR handoff comment 10676)
  • The transport incident that delayed recording is captured on #584/#672 with remediation owned by #685 (resolver contract) and #686 (duplicate-launch guardrail)

What is blocked:

  • Blocker classification: no blocker

Who/what acts next:

  • Next actor: merger
  • Required action: canonical merge workflow for PR #684, then #683 closure follows the merge per workflow
  • Do not do: do not merge without re-verifying live head equals d302602 and the verdict remains visible; do not close #683 before PR #684 lands on master; do not absorb #680/#604 work; do not use the preserved 300a4ca branch

Canonical Issue State

STATE:
pr-approve-verdict-recorded — PR #684 carries formal APPROVE at head d302602567a1fb7842fe5486fa5a49509506ec8b; awaiting merger workflow

WHO_IS_NEXT:
merger

NEXT_ACTION:
Run canonical merge workflow for PR #684; close #683 via the Closes reference when the merge lands

NEXT_PROMPT:

Role: MERGER
Repo: prgs / Scaled-Tech-Consulting / Gitea-Tools
PR: #684 | Issue: #683 | Head: d302602567a1fb7842fe5486fa5a49509506ec8b | Base: master

1. gitea_whoami + gitea_load_review_workflow + gitea_resolve_task_capability(merge_pr) on prgs-merger.
2. Verify live head still equals d302602 and the APPROVE verdict is visible at that head (gitea_get_pr_review_feedback: approval_at_current_head=true, has_blocking_change_requests=false).
3. Acquire/adopt merger lease; perform the merge via gitea_merge_pr; post-merge cleanup per workflow (branch delete, #683 closure path, root advance per runbook).
4. If the head moved or the verdict is not visible, stop and hand back to reviewer.

WHAT_HAPPENED:
Independent reviewer (sysadmin/prgs-reviewer, distinct from author jcwalker3/prgs-author) reviewed PR #684 at pinned head d302602 (evidence comment 10606: scope limited to #683, all eight acceptance criteria verified, rejected 300a4ca patterns proven absent, full-suite parity). The formal verdict recording was deferred through the #584/#672 transport incident and completed after operator reconnect of the client-managed gitea-reviewer namespace.

WHY:
#683's hardening ACs are implemented and verified; the remaining review findings are non-blocking follow-up candidates and do not gate this issue's PR.

RELATED_PRS:
PR #684 (Closes #683 — formal APPROVE at d302602); #680 separate landing sequence; #681 preserve branch remains evidence only

BLOCKERS:
none

VALIDATION:

  • Read-back after submission: approval_visible=true; approval_at_current_head=true; latest_approved_head_sha=d302602567a1fb7842fe5486fa5a49509506ec8b; has_blocking_change_requests=false; stale=false; dismissed=false
  • Provenance: client-managed gitea-reviewer namespace; identity sysadmin/prgs-reviewer via keychain; runtime parity 22698c1; restart_required=false; no GITEA_DUMMY process in inventory
  • Incident linkage: #685, #686, #584 (comments 10640/10650), #672 (comments 10644/10656)

LAST_UPDATED_BY:
sysadmin / prgs-reviewer / 2026-07-12

[THREAD STATE LEDGER] Issue #683 — PR #684 carries a formal APPROVE verdict at head d302602; merger workflow is next What is true now: - Server-side decision state: PR #684 (Closes #683) carries a formal APPROVE verdict by sysadmin/prgs-reviewer recorded via the review API; read-back shows approval_visible=true, approval_at_current_head=true at head d302602567a1fb7842fe5486fa5a49509506ec8b, has_blocking_change_requests=false - Local verdict/state: matches server; independent review evidence in PR #684 comment 10606 at the same head - Latest known validation: focused 19 passed; changed suites 30 passed; guard/lease/review/merge suites 155 passed; full suite 2615 passed / 6 skipped; compileall OK; four findings non-blocking What changed: - Formal verdict recorded in a fresh client-managed reviewer run after operator reconnect (fresh lease comment 10670, session 76849-c083d5c9e7da; verdict read back at 2026-07-12T18:53:34-05:00; PR handoff comment 10676) - The transport incident that delayed recording is captured on #584/#672 with remediation owned by #685 (resolver contract) and #686 (duplicate-launch guardrail) What is blocked: - Blocker classification: no blocker Who/what acts next: - Next actor: merger - Required action: canonical merge workflow for PR #684, then #683 closure follows the merge per workflow - Do not do: do not merge without re-verifying live head equals d302602 and the verdict remains visible; do not close #683 before PR #684 lands on master; do not absorb #680/#604 work; do not use the preserved 300a4ca branch ## Canonical Issue State STATE: pr-approve-verdict-recorded — PR #684 carries formal APPROVE at head d302602567a1fb7842fe5486fa5a49509506ec8b; awaiting merger workflow WHO_IS_NEXT: merger NEXT_ACTION: Run canonical merge workflow for PR #684; close #683 via the Closes reference when the merge lands NEXT_PROMPT: ```text Role: MERGER Repo: prgs / Scaled-Tech-Consulting / Gitea-Tools PR: #684 | Issue: #683 | Head: d302602567a1fb7842fe5486fa5a49509506ec8b | Base: master 1. gitea_whoami + gitea_load_review_workflow + gitea_resolve_task_capability(merge_pr) on prgs-merger. 2. Verify live head still equals d302602 and the APPROVE verdict is visible at that head (gitea_get_pr_review_feedback: approval_at_current_head=true, has_blocking_change_requests=false). 3. Acquire/adopt merger lease; perform the merge via gitea_merge_pr; post-merge cleanup per workflow (branch delete, #683 closure path, root advance per runbook). 4. If the head moved or the verdict is not visible, stop and hand back to reviewer. ``` WHAT_HAPPENED: Independent reviewer (sysadmin/prgs-reviewer, distinct from author jcwalker3/prgs-author) reviewed PR #684 at pinned head d302602 (evidence comment 10606: scope limited to #683, all eight acceptance criteria verified, rejected 300a4ca patterns proven absent, full-suite parity). The formal verdict recording was deferred through the #584/#672 transport incident and completed after operator reconnect of the client-managed gitea-reviewer namespace. WHY: #683's hardening ACs are implemented and verified; the remaining review findings are non-blocking follow-up candidates and do not gate this issue's PR. RELATED_PRS: PR #684 (Closes #683 — formal APPROVE at d302602); #680 separate landing sequence; #681 preserve branch remains evidence only BLOCKERS: none VALIDATION: - Read-back after submission: approval_visible=true; approval_at_current_head=true; latest_approved_head_sha=d302602567a1fb7842fe5486fa5a49509506ec8b; has_blocking_change_requests=false; stale=false; dismissed=false - Provenance: client-managed gitea-reviewer namespace; identity sysadmin/prgs-reviewer via keychain; runtime parity 22698c1; restart_required=false; no GITEA_DUMMY process in inventory - Incident linkage: #685, #686, #584 (comments 10640/10650), #672 (comments 10644/10656) LAST_UPDATED_BY: sysadmin / prgs-reviewer / 2026-07-12
sysadmin commented 2026-07-12 18:59:38 -05:00
Owner
Copy Link
Copy Source

Canonical Issue State

STATE: closed — satisfied by merged PR #684 on master

WHO_IS_NEXT: reconciler

NEXT_ACTION: Optional label/branch hygiene only: clear stale status:pr-open if still present on this closed issue; delete PR source branch if still on remote. No further product work on #683. Open defects #685 and #686 remain independent non-blocking follow-ups.

NEXT_PROMPT:

Role: RECONCILER (optional)
Issue #683 is closed via Closes from PR #684.
Merge commit: 56f1230a10a21b16226c792a5bcdc4398c97b09a
Approved head on master: d302602567a1fb7842fe5486fa5a49509506ec8b
Clear stale status:pr-open; delete fix/issue-683-workflow-guard-hardening if present.
Do not reopen #683 for #685/#686 work — file/keep those as separate issues.

WHAT_HAPPENED:
PR #684 (Closes #683) was merged by sysadmin/prgs-merger after independent formal APPROVE by sysadmin/prgs-reviewer at head d302602567 (evidence comment 10606; handoff comments 10676/10683).

Authorization: Explicit operator MERGER prompt with confirmation MERGE PR 684 and pinned expected head. Reviewer handoff alone was not treated as merge authorization.

Result

  • Issue state: closed
  • PR #684: merged
  • Merge commit: 56f1230a10a21b16226c792a5bcdc4398c97b09a
  • Approved head present on master (ancestor of prgs/master)
  • Merger lease adopted (10691) then post-merge moot lease released (10696)
  • Canonical PR handoff: comment 10698 on #684

Non-blocking follow-ups (not merge blockers, not absorbed)

  • #685 — resolver contract / config-touch defect
  • #686 — duplicate manual role-server launch guardrail

WHY: #683 acceptance criteria were implemented and independently approved; merge gates and operator confirmation were satisfied.

RELATED_PRS: PR #684 merged (Closes #683); #685 #686 remain open product defects

BLOCKERS: none for #683. Residual: possible stale status:pr-open label and undeleted source branch (reconciler; merger profile lacks branch.delete).

VALIDATION:

  • Issue #683 state=closed after merge
  • merge_commit_sha=56f1230a10a21b16226c792a5bcdc4398c97b09a on master
  • d302602 is ancestor of prgs/master

LAST_UPDATED_BY: sysadmin / prgs-merger / MERGER / 2026-07-12

## Canonical Issue State STATE: closed — satisfied by merged PR #684 on master WHO_IS_NEXT: reconciler NEXT_ACTION: Optional label/branch hygiene only: clear stale `status:pr-open` if still present on this closed issue; delete PR source branch if still on remote. No further product work on #683. Open defects #685 and #686 remain independent non-blocking follow-ups. NEXT_PROMPT: ```text Role: RECONCILER (optional) Issue #683 is closed via Closes from PR #684. Merge commit: 56f1230a10a21b16226c792a5bcdc4398c97b09a Approved head on master: d302602567a1fb7842fe5486fa5a49509506ec8b Clear stale status:pr-open; delete fix/issue-683-workflow-guard-hardening if present. Do not reopen #683 for #685/#686 work — file/keep those as separate issues. ``` WHAT_HAPPENED: PR #684 (Closes #683) was merged by sysadmin/prgs-merger after independent formal APPROVE by sysadmin/prgs-reviewer at head d302602567a1fb7842fe5486fa5a49509506ec8b (evidence comment 10606; handoff comments 10676/10683). **Authorization:** Explicit operator MERGER prompt with confirmation `MERGE PR 684` and pinned expected head. Reviewer handoff alone was not treated as merge authorization. **Result** - Issue state: closed - PR #684: merged - Merge commit: `56f1230a10a21b16226c792a5bcdc4398c97b09a` - Approved head present on master (ancestor of prgs/master) - Merger lease adopted (10691) then post-merge moot lease released (10696) - Canonical PR handoff: comment 10698 on #684 **Non-blocking follow-ups (not merge blockers, not absorbed)** - #685 — resolver contract / config-touch defect - #686 — duplicate manual role-server launch guardrail WHY: #683 acceptance criteria were implemented and independently approved; merge gates and operator confirmation were satisfied. RELATED_PRS: PR #684 merged (Closes #683); #685 #686 remain open product defects BLOCKERS: none for #683. Residual: possible stale `status:pr-open` label and undeleted source branch (reconciler; merger profile lacks branch.delete). VALIDATION: - Issue #683 state=closed after merge - merge_commit_sha=56f1230a10a21b16226c792a5bcdc4398c97b09a on master - d302602 is ancestor of prgs/master LAST_UPDATED_BY: sysadmin / prgs-merger / MERGER / 2026-07-12
sysadmin removed the status:pr-open label 2026-07-12 20:13:45 -05:00
sysadmin commented 2026-07-12 20:14:46 -05:00
Owner
Copy Link
Copy Source

Canonical Issue State

STATE:
reconciled — source branch cleanup handed off; issue closed; labels cleaned; residual remote source branch remains pending authorized delete capability

WHO_IS_NEXT:
user

NEXT_ACTION:
No further product work on #683. Operator configures a profile with gitea.branch.delete (or grants it to reconciler), reconnects client-managed namespace, then runs guarded gitea_cleanup_merged_pr_branch for PR #684. Do not use raw git branch deletion.

NEXT_PROMPT:

Role: RECONCILER (after gitea.branch.delete is configured) or USER/operator
Repo: prgs / Scaled-Tech-Consulting / Gitea-Tools
PR: #684 (merged) | branch: fix/issue-683-workflow-guard-hardening @ d302602567a1fb7842fe5486fa5a49509506ec8b

1. Fresh client-managed runtime; prove master parity (not stale).
2. Confirm PR #684 merged; merge_commit 56f1230a10a21b16226c792a5bcdc4398c97b09a; branch fully merged (ancestor of master).
3. Confirm no worktree/lease/canonical preservation requirement references this branch (preserve branch chore/issue-681-preserve-review-session-wip @ 300a4ca is UNRELATED — do not touch).
4. Use ONLY gitea_cleanup_merged_pr_branch(pr_number=684, confirmation=…, branch=fix/issue-683-workflow-guard-hardening) — never raw git push delete.
5. If capability still missing: configure a reconciler profile with gitea.branch.delete, reconnect, retry — do not improvise.

WHAT_HAPPENED:
Fresh RECONCILER session (sysadmin / prgs-reconciler) on client-managed gitea-reconciler after operator reconnect from the stale 22698c1 cohort.

Fresh runtime preflight (passed)

  • profile: prgs-reconciler
  • identity: sysadmin
  • client-managed namespace provenance: probe_source=client_namespace, ide_namespace_proven=true, healthy
  • startup_head = current_head = 56f1230a10a21b16226c792a5bcdc4398c97b09a
  • master_parity.stale=false, restart_required=false, in_parity=true
  • mutation gate not blocking for stale runtime
  • root checkout clean on master @ 56f1230a…
  • reconciliation workflow boundary: fresh session (no stale capability/workflow reuse from 22698c1 cohort); resolver only after freshness proven

Canonical state verification

  • PR #684: merged (closed); merge_commit_sha=56f1230a10a21b16226c792a5bcdc4398c97b09a
  • Source head d302602567a1fb7842fe5486fa5a49509506ec8b is ancestor of prgs/master (fully merged)
  • Issue #683: closed (proof: comment 10702 merger handoff; live state=closed)
  • PR comment 10698: merger canonical handoff to reconciler
  • Issue comment 10702: closed via Closes; residual label/branch hygiene named
  • Active lease: none (post-merge moot lease already released, comment 10696; cleanup tool: already released/terminal)
  • Active claim / terminal-review lock / assignment: none (assignee empty; issue-claim inventory empty; no live issue locks)
  • #685: open (status:ready)
  • #686: open (status:ready); comment 10712 adds per-client installer verification ACs
  • #672: open; comment 10707 records cross-client namespace discrepancy vs stale-runtime separation
  • Preservation branch chore/issue-681-preserve-review-session-wip untouched at 300a4ca08fee3a195efa723d17eb6c64a3ef7a4f

Label reconciliation

  • Before: labels included stale status:pr-open
  • Action: gitea_set_issue_labels after capability resolve — removed only status:pr-open
  • After (verified): anti-stomp, preflight, safety, type:guardrail, workflow-hardening — no status:pr-open

Source-branch handling

  • Remote branch still present: fix/issue-683-workflow-guard-hardening @ d302602567a1fb7842fe5486fa5a49509506ec8b
  • Fully merged through PR #684 (ancestor of master)
  • gitea_resolve_task_capability(cleanup_merged_pr_branch): required permission gitea.branch.delete, role reconciler; no profile configured with that permission (matching_configured_profile: [])
  • Active profile allowed ops: read, issue.comment, pr.comment, pr.close only
  • No raw git branch deletion performed
  • Handoff: authorized role/profile with gitea.branch.delete → gitea_cleanup_merged_pr_branch only

Cross-client notes (not conflated)

  • One client: author + reviewer only
  • This client: callable reconciler; prior operational issue here was stale runtime after master advanced, not missing namespace
  • Recorded on #672 (comment 10707) and #686 (comment 10712)

WHY:
#683 product work is complete and closed via merged PR #684. Reconciler completed all authorized hygiene (labels + durable final record). Residual branch delete requires a permission no configured profile currently has.

RELATED_PRS:
PR #684 merged @ 56f1230a10a21b16226c792a5bcdc4398c97b09a; #685 #686 remain open; preserve branch evidence only

BLOCKERS:
none for #683 product closure. Residual hygiene only: remote source branch delete capability missing (gitea.branch.delete not configured on any profile). Unblock residual branch cleanup when: operator configures gitea.branch.delete on a reconciler (or authorized) profile, reconnects client-managed namespace, and runs gitea_cleanup_merged_pr_branch for PR #684.

VALIDATION:

  • Live PR #684 merge_commit_sha = 56f1230a10
  • Live issue #683 state=closed; status:pr-open absent
  • git ls-remote: source branch still at d302602; preserve branch still at 300a4ca
  • lease: no active; post-merge cleanup already terminal
  • root clean master @ 56f1230a

LAST_UPDATED_BY:
sysadmin / prgs-reconciler / RECONCILER / 2026-07-12

## Canonical Issue State STATE: reconciled — source branch cleanup handed off; issue closed; labels cleaned; residual remote source branch remains pending authorized delete capability WHO_IS_NEXT: user NEXT_ACTION: No further product work on #683. Operator configures a profile with `gitea.branch.delete` (or grants it to reconciler), reconnects client-managed namespace, then runs guarded `gitea_cleanup_merged_pr_branch` for PR #684. Do not use raw git branch deletion. NEXT_PROMPT: ```text Role: RECONCILER (after gitea.branch.delete is configured) or USER/operator Repo: prgs / Scaled-Tech-Consulting / Gitea-Tools PR: #684 (merged) | branch: fix/issue-683-workflow-guard-hardening @ d302602567a1fb7842fe5486fa5a49509506ec8b 1. Fresh client-managed runtime; prove master parity (not stale). 2. Confirm PR #684 merged; merge_commit 56f1230a10a21b16226c792a5bcdc4398c97b09a; branch fully merged (ancestor of master). 3. Confirm no worktree/lease/canonical preservation requirement references this branch (preserve branch chore/issue-681-preserve-review-session-wip @ 300a4ca is UNRELATED — do not touch). 4. Use ONLY gitea_cleanup_merged_pr_branch(pr_number=684, confirmation=…, branch=fix/issue-683-workflow-guard-hardening) — never raw git push delete. 5. If capability still missing: configure a reconciler profile with gitea.branch.delete, reconnect, retry — do not improvise. ``` WHAT_HAPPENED: Fresh RECONCILER session (sysadmin / prgs-reconciler) on client-managed `gitea-reconciler` after operator reconnect from the stale `22698c1` cohort. ### Fresh runtime preflight (passed) - profile: `prgs-reconciler` - identity: `sysadmin` - client-managed namespace provenance: `probe_source=client_namespace`, `ide_namespace_proven=true`, healthy - `startup_head` = `current_head` = `56f1230a10a21b16226c792a5bcdc4398c97b09a` - `master_parity.stale=false`, `restart_required=false`, `in_parity=true` - mutation gate not blocking for stale runtime - root checkout clean on `master` @ `56f1230a…` - reconciliation workflow boundary: fresh session (no stale capability/workflow reuse from 22698c1 cohort); resolver only after freshness proven ### Canonical state verification - PR #684: **merged** (closed); `merge_commit_sha=56f1230a10a21b16226c792a5bcdc4398c97b09a` - Source head `d302602567a1fb7842fe5486fa5a49509506ec8b` is ancestor of prgs/master (fully merged) - Issue #683: **closed** (proof: comment 10702 merger handoff; live state=closed) - PR comment 10698: merger canonical handoff to reconciler - Issue comment 10702: closed via Closes; residual label/branch hygiene named - Active lease: **none** (post-merge moot lease already released, comment 10696; cleanup tool: already released/terminal) - Active claim / terminal-review lock / assignment: **none** (assignee empty; issue-claim inventory empty; no live issue locks) - #685: **open** (status:ready) - #686: **open** (status:ready); comment 10712 adds per-client installer verification ACs - #672: **open**; comment 10707 records cross-client namespace discrepancy vs stale-runtime separation - Preservation branch `chore/issue-681-preserve-review-session-wip` **untouched** at `300a4ca08fee3a195efa723d17eb6c64a3ef7a4f` ### Label reconciliation - Before: labels included stale `status:pr-open` - Action: `gitea_set_issue_labels` after capability resolve — removed **only** `status:pr-open` - After (verified): `anti-stomp`, `preflight`, `safety`, `type:guardrail`, `workflow-hardening` — **no** `status:pr-open` ### Source-branch handling - Remote branch still present: `fix/issue-683-workflow-guard-hardening` @ `d302602567a1fb7842fe5486fa5a49509506ec8b` - Fully merged through PR #684 (ancestor of master) - `gitea_resolve_task_capability(cleanup_merged_pr_branch)`: required permission `gitea.branch.delete`, role reconciler; **no profile configured** with that permission (`matching_configured_profile: []`) - Active profile allowed ops: read, issue.comment, pr.comment, pr.close only - **No raw git branch deletion performed** - **Handoff:** authorized role/profile with `gitea.branch.delete` → `gitea_cleanup_merged_pr_branch` only ### Cross-client notes (not conflated) - One client: author + reviewer only - This client: callable reconciler; prior operational issue here was stale runtime after master advanced, not missing namespace - Recorded on #672 (comment 10707) and #686 (comment 10712) WHY: #683 product work is complete and closed via merged PR #684. Reconciler completed all authorized hygiene (labels + durable final record). Residual branch delete requires a permission no configured profile currently has. RELATED_PRS: PR #684 merged @ `56f1230a10a21b16226c792a5bcdc4398c97b09a`; #685 #686 remain open; preserve branch evidence only BLOCKERS: none for #683 product closure. Residual hygiene only: remote source branch delete capability missing (`gitea.branch.delete` not configured on any profile). Unblock residual branch cleanup when: operator configures `gitea.branch.delete` on a reconciler (or authorized) profile, reconnects client-managed namespace, and runs `gitea_cleanup_merged_pr_branch` for PR #684. VALIDATION: - Live PR #684 merge_commit_sha = 56f1230a10a21b16226c792a5bcdc4398c97b09a - Live issue #683 state=closed; status:pr-open absent - git ls-remote: source branch still at d302602; preserve branch still at 300a4ca - lease: no active; post-merge cleanup already terminal - root clean master @ 56f1230a LAST_UPDATED_BY: sysadmin / prgs-reconciler / RECONCILER / 2026-07-12
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
allocator
Controller-owned work allocator
 anti-stomp
Prevent concurrent LLM session stomping
 architecture
Architecture / structural design
 bug
chore
codex
OpenAI Codex client / workflow session surface
 concurrency
Concurrent session safety
 contamination
Workflow or session contamination incident
 control-plane
MCP control-plane coordination and allocation authority
 dashboard
MCP operational dashboard/queue view
 database
Internal coordination storage (SQLite/Postgres)
 design
Design / investigation, no implementation
 documentation
Docs / runbooks
 enhancement
New feature or improvement
 gitea
Gitea MCP workflow
 glitchtip
GlitchTip integration
 important
incident
Operational or process incident requiring durable audit trail
 incident-bridge
Sentry-to-Gitea incident bridging
 integration
Integration testing
 jenkins
Jenkins integration
 labels
Label taxonomy management
 leases
Lease adopt/release/expire lifecycle
 mcp
MCP server / tooling
 mcp-health
MCP namespace and runtime health
 mcp-menu
MCP menu surface
 multi-project
Work spanning multiple monitoring projects or Gitea repos
 mutating
Mutating action; requires gating
 nice-to-have
observability
Observability, metrics, traces, error reporting
 portability
Cross-platform / portability
 preflight
Shared preflight gates before mutation
 protected-branch
Protected branch / stable-branch policy concern
 queue
Work queue visibility and allocation
 read-only
Read-only, no mutation
 reconnect
MCP client reconnect/reload recovery path
 recovery
Recovery paths for stale/foreign leases
 refactor
Code refactor / restructure
 release
Release / versioning
 reliability
Reliability / failure handling
 resumable-review
Persist and resume prepared review verdicts across sessions
 reviewer
Reviewer workflow tooling
 roadmap
Roadmap / umbrella issue
 safety
Safety rails and fail-closed mutation guards
 security
Security / trust boundary
 self-hosted
Self-hosted infrastructure integration
 sentry
Sentry error monitoring integration
 stale-runtime
Stale backend daemon / runtime-vs-master parity failures
 status:blocked
Issue is blocked
 status:in-progress
Issue is being worked on
 status:pr-open
Issue has an open pull request
 status:ready
Issue is ready for work
 terminal-lock
Terminal review lock (#332) path
 testing
Tests / test coverage
 tracker
Issue tracker hygiene / meta
 type:bug
Bug or defect
 type:feature
Feature or enhancement
 type:feature
Feature or enhancement
 type:guardrail
Safety gate or guardrail
 visibility
Workflow state visibility for LLMs/operators
 workflow
Cross-tool workflow
 workflow-hardening
LLM workflow coordination hardening
 workflow-hardening
LLM workflow coordination hardening
No labels anti-stomp preflight safety type:guardrail workflow-hardening
Milestone
No items
No Milestone
Projects
Clear projects
No projects
Assignees
Clear assignees
jcwalker3 sysadmin
No Assignees
2 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Scaled-Tech-Consulting/Gitea-Tools#683