More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    🩺 CI Doctor

    2

    Diagnose CI step failures and jobs that fail before steps run. Trace empty-step states and matrix cancellations to evidence, then propose the smallest reviewable fix.

    $12

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    🩺 CI Doctor

    🩺 CI Doctor

    Example session with this skill installed

    Use $ci-doctor to diagnose why GitHub Actions job 123456789 failed with steps: []. Tell me the next safe check, but do not change files or rerun CI.

    • Read your context and instructions
    • Compiled the ci doctor

    Run State: GitHub Actions / initialize; conclusion: failure; normal steps: none. Root Cause: unresolved pre-step failure. Evidence: the job exposes an empty steps list, no runner assignment, and no decisive annotation. Fix: inspect the workflow at the run SHA and the check-run annotations. Confidence: low until those artifacts are available. Approval Boundary: no code, secret, or workflow change is justified yet.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Diagnose the failure before editing CI

    CI Doctor turns a failed pipeline into one evidence-backed diagnosis, a bounded fix, and a clear approval boundary. It starts with GitHub Actions and recognizes GitLab CI and CircleCI markers from supplied configuration and logs.

    Handles failures with no steps or logs

    When a job exposes no normal steps or an empty steps array, CI Doctor does not guess that application code or a secret failed. It inspects run and job state, annotations, the workflow at the failing commit, runner assignment and labels, environment protections, job conditions and dependencies, reusable workflows, and matrix siblings. If the evidence remains incomplete, it reports an unresolved pre-step failure and names the next read-only check.

    Understands matrix behavior

    CI Doctor traces fail-fast cancellations back to the earliest required sibling failure and distinguishes job-level allowed failures from step-level continue-on-error behavior. It preserves intentional experimental jobs rather than turning required failures green.

    Evidence first, mutations by approval

    • Inspects workflow YAML, manifests, lockfiles, tests, diffs, logs, and run metadata.
    • Separates the primary cause from cascading errors and cancelled matrix siblings.
    • Returns run state, root cause or explicit uncertainty, short evidence, the smallest fix or next check, and confidence and unknowns.
    • Never treats a secret reference or empty step list as proof that a secret is missing.
    • Asks before editing files, rerunning CI, setting secrets, committing, pushing, or opening a pull request.

    Live private-repository inspection requires authenticated GitHub access; otherwise provide workflow files and redacted run, job, annotation, or log data. An empty steps array can narrow the evidence path but may not expose a final cause, so CI Doctor may report an unresolved pre-step failure. GitLab CI and CircleCI are diagnosed from supplied config and logs. It cannot reveal secret inventory, repair remote runner infrastructure or GitHub UI settings, or mutate state without confirmation.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 24 days ago

    Requires

    GitHub
    Python
    Git

    Permissions

    Read Files
    Write Files
    Network Access

    Allowed Hosts

    api.github.com

    File Scopes

    .github/workflows/**/*.yml
    .github/workflows/**/*.yaml
    package.json
    requirements.txt
    go.mod
    Cargo.toml
    pom.xml

    Shell and network access support read-only GitHub CLI and git inspection of run and job metadata, annotations, workflow revisions, and logs. Read access covers workflow files, manifests, lockfiles, tests, and relevant diffs. Write access is used only after explicit approval for a proposed fix. CI Doctor asks again before rerunning CI, setting secrets, committing, pushing, or opening a pull request. It does not request or print secret values. Network access is limited to GitHub through the authenticated CLI or api.github.com.

    Listed5 months ago
    Updated24 days ago

    What's inside

    Frequently Asked Questions

    It treats an empty or missing steps list as missing execution evidence, not proof that application code or a secret failed. It checks run and job state, annotations, the workflow at the run SHA, runner labels, environment protections, job conditions, reusable workflows, and matrix siblings before proposing a cause.

    GitHub Actions has the deepest live-run command and schema coverage. CI Doctor also recognizes GitLab CI and CircleCI and diagnoses them from the configuration and redacted logs you provide.

    For live GitHub inspection, the agent needs authenticated GitHub CLI access to the repository. You can instead supply workflow files plus redacted run, job, annotation, or log data. Secret values are never required.

    No. Inspection starts read-only. It asks before editing files, rerunning CI, setting secrets, committing, pushing, or opening a pull request.

    A five-file SKILL.md bundle with the core diagnostic workflow, 60 failure and state patterns, cautious fix templates, a GitHub Actions schema reference, and the agent interface metadata.

    It traces fail-fast cancellations to the earliest required sibling failure. It distinguishes job-level allowed failures from step-level continue-on-error behavior and preserves intentional experimental jobs unless repository policy says otherwise.

    New in DevOps & Cloud