fix(maintainer): accept retryable E2E history

[cv]

Summary

The maintainer merge-gate checker now accepts immutable E2E retry history under the same fail-closed rule already enforced by the controller and required observer. This unblocks a later successful check only when every older exact-identity check is a completed failure with a valid supported retry marker.

Related Issue

Unblocks #7582 and the dependent #7557 release sequence.

Changes

• Select the highest-ID exact coordination check after validating the full retry history.
• Reject malformed markers, unsupported reasons, never-retry titles, active history, identity drift, and foreign GitHub Apps.
• Add a focused regression suite for all supported retry reasons and fail-closed histories.

Type of Change

• Code change (feature, bug fix, or refactor)
• Code change with doc updates
• Doc only (prose changes, no code sample modifications)
• Doc only (includes code sample changes)

Quality Gates

• Tests added or updated for changed behavior
• Existing tests cover changed behavior — justification:
• Tests not applicable — justification:
• Docs updated for user-facing behavior changes
• Docs not applicable — justification: the canonical maintainer guide and E2E runbook already document this exact retry-history contract; this change aligns the checker implementation.
• Sensitive paths changed (security, policy, credentials, preflight, onboarding, inference, runner, sandbox, or messaging)
• Sensitive-path review completed or maintainer-approved waiver recorded — reviewer/approval link/justification: independent exact-diff review by Codex Desktop passed; the parser and selector mirror the existing controller and observer allowlist and fail-closed rules.
• Non-success, skipped, or missing CI check accepted by maintainer — check name, approval link, and follow-up issue:

Documentation Writer Review

• Documentation writer subagent reviewed the completed changes
• Result: no-docs-needed
• Evidence: Existing .agents/skills/nemoclaw-maintainer-day/MERGE-GATE.md and test/e2e/README.md already define the supported reasons, immutable history, highest-ID selection, and fail-closed conditions.
• Agent: Codex Desktop

DGX Station Hardware Evidence

• Tested on DGX Station
• Tested commit: Not applicable
• Station profile/scenario: Not applicable
• Result: Not applicable
• Supporting evidence: Not applicable

Verification

• PR description includes a Signed-off-by: line and every commit appears as Verified in GitHub
• Normal pre-commit, commit-msg, and pre-push hooks passed, or npm run check:diff passed when hooks were skipped or unavailable
• Targeted behavior tests pass for the current change set, or tests are marked not applicable above — npx vitest run --project integration test/skills/check-gates-evidence.test.ts test/skills/check-gates-retry-history.test.ts test/skills/check-gates-fork-evidence.test.ts test/skills/check-gates-compliance.test.ts passed 4 files and 160 tests; npm run test-size:check passed.
• Applicable broad gate passed — npm test for broad runtime/test-harness changes; npm run check for repo-wide validation/coverage changes — not applicable to the two-file internal maintainer-checker change; the complete checker suite and normal hooks passed.
• Quality Gates section completed with required justifications or waivers
• No secrets, API keys, or credentials committed
• npm run docs builds without warnings (doc changes only)
• Doc pages follow the style guide (doc changes only)
• New doc pages include SPDX header and frontmatter (new pages only)

---

Signed-off-by: Carlos Villela cvillela@nvidia.com

Summary by CodeRabbit

• Bug Fixes

  • Improved E2E merge-gate validation for retryable failures and retry history.
  • Correctly recognizes supported retry markers and excludes non-retryable runs.
  • More reliably identifies the current E2E coordination check when multiple checks exist.
  • Fails closed when retry evidence is incomplete, malformed, unrelated, or includes active or unsupported checks.

• Tests

  • Added coverage for successful retries following supported failures.
  • Added validation for incomplete, invalid, and non-retryable retry histories.

[coderabbitai]

📝 Walkthrough

Walkthrough

The E2E coordination gate now recognizes retryable failure markers, excludes non-retryable titles, selects the current coordination check from ordered history, and validates multiple candidate checks. Tests cover accepted retry histories and fail-closed scenarios.

Changes

E2E retry-history validation

Layer / File(s)	Summary
Retry classification and current-run selection .agents/skills/nemoclaw-maintainer-day/scripts/check-gates.ts	Defines retry marker and failure classifications, parses completed failure summaries, excludes never-retry titles, and selects the current coordination check from ordered runs.
Gate evidence integration and retry-history coverage .agents/skills/nemoclaw-maintainer-day/scripts/check-gates.ts, test/skills/check-gates-retry-history.test.ts	Validates all matching coordination candidates before selecting one current run, with tests for accepted retry histories and invalid, active, mismatched, or unsupported histories.

Estimated code review effort: 3 (Moderate) | ~20 minutes

Suggested labels: area: e2e

Suggested reviewers: prekshivyas

Sequence Diagram(s)

sequenceDiagram
  participant GateEvaluation
  participant GitHubChecks
  participant fetchE2eCoordinationEvidence
  GateEvaluation->>GitHubChecks: retrieve E2E coordination checks
  GitHubChecks->>fetchE2eCoordinationEvidence: return candidate runs
  fetchE2eCoordinationEvidence->>fetchE2eCoordinationEvidence: validate candidates and retry history
  fetchE2eCoordinationEvidence-->>GateEvaluation: return current coordination evidence

Loading

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly summarizes the main change: accepting retryable E2E history in the maintainer gate.

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch codex/check-gates-retry-history

---

_{Comment @coderabbitai help to get the list of available commands.}

[github-code-quality]

Code Coverage Overview

Languages: TypeScript

TypeScript / code-coverage/plugin

The overall coverage in commit 7cd0ff7 in the codex/check-gates-re... branch remains at 96%, unchanged from commit d4a859a in the main branch.

TypeScript / code-coverage/cli

The overall coverage in commit 7cd0ff7 in the codex/check-gates-re... branch remains at 81%, unchanged from commit d4a859a in the main branch.

Show a code coverage summary of the most impacted files.

File	main d4a859a	codex/check-gates-re... 7cd0ff7	+/-
src/lib/platform.ts	89%	84%	-5%
src/lib/messagi...nnels/policy.ts	100%	100%	0%
src/lib/sandbox...rce-identity.ts	88%	88%	0%
src/lib/onboard...e-patch-flow.ts	71%	76%	+5%

---

_{Updated July 26, 2026 12:43 UTC}

[github-actions]

PR Review Advisor — No blocking findings reported

Advisor assessment: No blocking advisor findings reported
Next action: Review the warnings below.
Findings: 0 blockers · 1 warning · 0 suggestions

Model lanes

• GPT-5.6 Terra (primary): Completed · high confidence · 0 blockers · 1 warning · 0 suggestions
• Nemotron 3 Ultra (second opinion): Completed · high confidence · 0 blockers · 0 warnings · 0 suggestions
• Model comparison: normalized findings differ; normalized E2E selections match; Nemotron reported the same number of blockers, 1 fewer warning, the same number of suggestions.

Nemotron output stays in workflow artifacts and does not change the assessment above.

E2E guidance

Advisory only. E2E / PR Gate selects and runs jobs independently.

Recommended E2E: None

1 warning · 0 suggestions

Warnings

Warnings do not block.

PRA-1 Warning — Cover retry history for the legacy E2E check fallback

• Location: test/skills/check-gates-retry-history.test.ts:37
• Category: tests
• Problem: The changed selector also accepts marker-backed retry history for the legacy `E2E / PR Gate` fallback when no coordination-name check exists. The new tests exercise multi-attempt history only for `E2E / PR Gate Coordination`; existing evidence tests cover only one successful legacy check.
• Impact: A regression in legacy fallback history selection could make the merge gate reject valid retry evidence or accept an invalid legacy history without a focused test detecting it.
• Recommendation: Add a legacy-only fixture with an empty coordination endpoint, an older exact marker-backed legacy failure, and a later successful legacy check. Assert that the CI gate passes.
• Verification: Read `fetchE2eCoordinationEvidence` and confirm that it passes `E2E / PR Gate` checks to `currentE2eCoordinationCheck` when no coordination-name check exists.
• Test coverage: A `runGate` test that returns no coordination-name checks and two legacy checks: an older supported retry marker failure and a later success, asserting `gates.ci.pass` is true.
• Evidence: `.agents/skills/nemoclaw-maintainer-day/scripts/check-gates.ts` selects legacy claimed checks when `currentNameChecks.length === 0`, then calls `currentE2eCoordinationCheck(exactChecks)`. `test/skills/check-gates-retry-history.test.ts` builds all retry-history fixtures with `coordinationCheck`, whose default name is `E2E / PR Gate Coordination`. `test/skills/check-gates-evidence.test.ts` has legacy-name coverage for a single successful check, not multi-attempt retry history.

Workflow run details

This automated review informs maintainers. Warnings and suggestions do not require a response. A maintainer decides whether to merge.