AgentFlows
ClaudeCI failureNot executed· 1 caveat

Read failing CI logs and suggest a fix

workflow_run: completed (failure)

`additional_permissions` 'Currently supports actions: read' (usage.md).

Estimated cost per run

~$0.35

Estimated, not measured.

Basis. ASSUMPTION (8 runner min x $0.006/min GitHub-hosted Linux 2-core (docs.github.com/en/billing/reference/actions-runner-pricing)) + ASSUMED 120,000 input / 6,000 output tokens at Claude Sonnet 5 API $2/MTok in, $10/MTok out (platform.claude.com/docs/en/about-claude/pricing). Token counts are not measured; replace after a real test run.

Verification

Last checked Oct 16, 2026.

inputs checked against source docs; workflow not executed.

Source

Open caveats

  • Linking a workflow_run back to its PR number needs extra steps; prompt assumes Claude can find it via gh. Untested.

Workflow YAML

Copy the file as-is. Highlighting is only for reading.

workflow.yml

name: CI failure helperon:  workflow_run:    workflows: ["CI"]    types: [completed]jobs:  diagnose:    if: github.event.workflow_run.conclusion == 'failure'    runs-on: ubuntu-latest    permissions:      contents: read      actions: read      pull-requests: write      id-token: write    steps:      - uses: actions/checkout@v6      - uses: anthropics/claude-code-action@v1        with:          anthropic_api_key: ${{ secrets.ANTHROPIC_API_KEY }}          additional_permissions: |            actions: read          prompt: |            REPO: ${{ github.repository }}            RUN ID: ${{ github.event.workflow_run.id }}            Read the failing job logs, find the root cause, and comment a fix on the related PR.

More workflows

  • Claude01

    Claude auto PR review

    Copied from the official Solutions Guide basic example (prompt shortened).

    pull_request: opened, synchronize

    Estimated cost~$0.18

    Not executed

    Checked Oct 16, 2026

  • Claude02

    Claude PR review with progress tracking

    `track_progress: true` posts a checkbox tracking comment (per Solutions Guide).

    pull_request: opened, synchronize, ready_for_review, reopened

    Estimated cost~$0.22

    Not executed

    Checked Oct 16, 2026

  • Claude03

    @claude mention assistant on issues and PRs

    No `prompt` → the action runs in tag/interactive mode and responds to the comment. `trigger_phrase` default is `@claude` (usage.md).

    issue_comment / pull_request_review_comment containing @claude

    Estimated cost~$0.27

    Not executed· 1 caveat

    Checked Oct 16, 2026