Skip to content

test: remove orphaned truth-gate fixture left by #1431 - #1475

Merged
groupthinking merged 1 commit into
mainfrom
claude/clever-heisenberg-dlp5ue
Aug 29, 2026
Merged

test: remove orphaned truth-gate fixture left by #1431#1475
groupthinking merged 1 commit into
mainfrom
claude/clever-heisenberg-dlp5ue

Conversation

@groupthinking

@groupthinking groupthinking commented Aug 7, 2026

Copy link
Copy Markdown
Owner

Canonical issue

Closes #1474

Outcome

tests/fixtures/agent_completion/ no longer exists. ci: retire the agent-completion truth gate (#1431) (8cd4a10ca) deleted both readers of pr_813.json but left the fixture in the tree.

Path State on main
scripts/ci/agent_completion_gate.py deleted by #1431
tests/unit/test_agent_completion_gate.py deleted by #1431
tests/fixtures/agent_completion/pr_813.json still present, no reader

The fixture encodes the retired gate's payload schema (policy.applicable, agent_login, run_id, head_sha). With nothing loading it, it reads as a live contract for a mechanism that no longer exists — the same misleading-artifact class #1431 and #1434 were closing.

Scope

  • Included: deletion of tests/fixtures/agent_completion/pr_813.json (the directory's only file).
  • Explicitly excluded:
    • docs/runbook/pr-remediation-report-2026-08-04.md:11 — the one remaining agent_completion mention in the repo. It is a historical runbook narrating the retirement, not a load of the fixture, and is left unchanged deliberately.
    • The truth-gate decision itself, which ci: retire the agent-completion truth gate #1431 settled.

Risk

  • Risk level: low
  • Failure mode: none identified. A deletion can only break a reader, and the orphan check below is exhaustive over the file types that could load it.
  • Rollback: git revert. No runtime, deployed, or data surface.

Verification

Head 2994e246, base 8c93de8bb. One file, -76.

  • Full unit suite passes with the fixture deleted — the decisive check, since a deletion can only fail by breaking a reader:

    $ pytest tests/unit -q --no-cov -n 4
    8073 passed, 5 xpassed in 61.93s
    

    Zero failures, zero collection errors. This also independently clears main after the 14-PR merge sweep of 2026-08-07 (d9bab00e8c93de8b), whose own post-merge CI was still queued when this was run.

  • Orphan status established by exhaustive grep, not assumed:

    $ grep -rn "agent_completion" --include=*.py --include=*.yml --include=*.yaml \
        --include=*.md --include=*.json --include=*.sh . \
        | grep -v '^./tests/fixtures/agent_completion/'
    ./docs/runbook/pr-remediation-report-2026-08-04.md:11:> `agent-completion-enforcement.yml`, `agent_completion_gate.py`, and
    

    Exactly one hit, and it is prose. No Python, workflow, or shell reference remains.

  • Both readers confirmed gone from main:

    $ git ls-tree -r --name-only origin/main | grep -E 'agent_completion_gate|agent-completion-enforcement'
    (no output)
    
  • The four tests that read tests/fixtures/ do not touch it. test_transcript_action_workflow.py, test_gemini_grok_failover.py, test_firestore_state.py, test_backend_main.py — a grep for agent_completion across all four returns nothing.

  • Required CI — queued. Every GitHub Actions job on this head was still queued 90 s after open; the sweep's 14 pushes saturated the runner pool. Only Vercel's own checks (which do not use GitHub runners) had reported, both success.

  • Review threads resolved — none open.

Note on the CodeRabbit status

Review skipped: excluded by label configuration, written ~3 s after this PR opened, while the label job was still queued — so no labels had been applied yet. This is confounded and should not be cited as evidence for #1424: this PR is also a draft, and .coderabbit.yaml sets auto_review.drafts: false, which is independently sufficient to explain the skip. #1440 remains the unconfounded case.

Production evidence

Not applicable. This deletes a test fixture; it ships no runtime code and has no deployable surface. Nothing under apps/web/**, so MERGE_POLICY.md gate 4 (preview) does not apply. The Vercel preview on this head is READY regardless.

Agent handoff

Agent provenance

Agent-authored under the PR remediation runbook. Found while verifying that #1409 ("arm the truth gate only on a real dispatch contract") is obsolete — that PR patches docs/agent-completion-truth-gate.md and tests/unit/test_agent_completion_gate.py, both now deleted from main. This fixture was the remaining artifact of the same retirement.

`ci: retire the agent-completion truth gate (#1431)` (8cd4a10) deleted
both readers of tests/fixtures/agent_completion/pr_813.json --
scripts/ci/agent_completion_gate.py and
tests/unit/test_agent_completion_gate.py -- but left the fixture itself
in the tree.

The fixture encodes the retired gate's payload schema (policy.applicable,
agent_login, run_id, head_sha). With no reader it reads as a live contract
for a mechanism that no longer exists, which is the same misleading-artifact
class #1431 and #1434 were closing.

Verified orphaned: a repo-wide grep for `agent_completion` outside the
fixture directory returns exactly one hit, a historical runbook narrating
the retirement (prose, not a load) -- left unchanged. None of the four
tests that read from tests/fixtures/ reference it.

Closes #1474

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BqvCiGTa8yXRhSsGdHWDgr
@vercel

vercel Bot commented Aug 7, 2026

Copy link
Copy Markdown
Contributor

The latest updates on your projects. Learn more about Vercel for GitHub.

Project Deployment Actions Updated (UTC)
v0-uvai Ready Ready Preview, v0 Aug 7, 2026 8:56pm

@coderabbitai

coderabbitai Bot commented Aug 7, 2026

Copy link
Copy Markdown
Contributor

Important

Review skipped

Auto reviews are limited based on label configuration.

🏷️ Required labels (at least one) (1)
  • [‘architecture-gap’, ‘bug’, ‘ci-cd’, ‘ci/cd’, ‘copilot-rabbit’, ‘documentation’, ‘duplicate’, ‘enhancement’, ‘frontend’, ‘github_actions’, ‘good first issue’, ‘help wanted’, ‘high-priority’, ‘invalid’, ‘javascript’, ‘ml-model’, ‘needs-triage’, ‘pipeline-critical’, ‘placeholder-code’, ‘priority:high’, ‘python’, ‘python:uv’, ‘question’, ‘styling’, ‘tests’, ‘v0’]

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository YAML (base), Repository UI (inherited), Organization UI (inherited)

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: b2d36b50-f54b-4d1d-befd-307675c30a24

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

  • 🔍 Trigger review

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

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

@github-actions

github-actions Bot commented Aug 7, 2026

Copy link
Copy Markdown

Dependency Review

✅ No vulnerabilities or license issues or OpenSSF Scorecard issues found.

Snapshot Warnings

⚠️: No snapshots were found for the head SHA 2994e24.
Ensure that dependencies are being submitted on PR branches and consider enabling retry-on-snapshot-warnings. See the documentation for more information and troubleshooting advice.

Scanned Files

None

@groupthinking groupthinking self-assigned this Aug 13, 2026
@groupthinking
groupthinking marked this pull request as ready for review August 26, 2026 23:32
@groupthinking
groupthinking merged commit 5473bcc into main Aug 29, 2026
34 of 35 checks passed
@groupthinking
groupthinking deleted the claude/clever-heisenberg-dlp5ue branch August 29, 2026 06:25
@linear-code

linear-code Bot commented Aug 29, 2026

Copy link
Copy Markdown
Contributor

GRV-437

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Orphaned truth-gate fixture left behind by #1431

2 participants