Best fit
A job can fail or wait before normal steps appear, leaving the owner unsure whether to inspect runners, policy, platform state, or workflow code.
// DevOps & Deployment
CI Failure Snapshot
Classify one redacted GitHub Actions failure and get cited evidence plus three bounded next checks before changing workflow code.
Format: Downloadable agent skill, not a browser app. You need an AI agent that can load its instructions and use the required tools.
Requirements: GitHub Actions run and job JSON exports Python 3.9 or newer on Windows, macOS, or Linux Local files only; no account connection required
You get: Bounded failure-stage classification Exact supplied fields used as evidence
Licence and purchase details
Review the package licence at the destination.
Review the offer and download options at the linked marketplace. Buying a skill does not include an AI-agent subscription or setup service.
What problem does CI Failure Snapshot solve?
A job can fail or wait before normal steps appear, leaving the owner unsure whether to inspect runners, policy, platform state, or workflow code.
Use it to
- Triage an empty-steps GitHub Actions job
- Separate runner wait from completed failure
- Catch contradictory run and job exports
- Choose the next evidence check before editing YAML
What it returns
- Bounded failure-stage classification
- Exact supplied fields used as evidence
- Three next read-only checks
- Confidence and explicit unknowns
A representative input and result
input Run conclusion failure; job conclusion failure; steps is empty; no runner name or decisive annotation is supplied.
result Classification: PRE_STEP_FAILURE. Confidence: medium. Inspect annotations, the workflow at the run SHA, runner labels, job conditions, dependencies, reusable workflows, and matrix siblings before changing code.
Access and approval boundaries
- Read access only to the user-supplied local input files.
- Write access only to the selected local output directory.
- No browser, network, credential, account, or environment-variable access is required.
Known limitations
- An empty steps array is missing execution evidence, not proof that a secret or application code failed.
- The snapshot does not inspect a repository, connect to GitHub, rerun CI, or change workflow files.
- Provider metadata can remain inconclusive and is then labeled unresolved.
// choose with context
Is CI Failure Snapshot the right skill?
It returns
Bounded failure-stage classification. Exact supplied fields used as evidence.
Do not use it as
An empty steps array is missing execution evidence, not proof that a secret or application code failed.
Compatibility, access, version, and licence
- Reads or accesses
- Declared local scope: run.json, job.json, ci-snapshot/**.
- Compatibility
- GitHub Actions run and job JSON exports Python 3.9 or newer on Windows, macOS, or Linux Local files only; no account connection required
- Version
- 1.0.0
- Licence
- Review the package licence and seller terms at the linked destination.
Questions
Does it connect to GitHub?
No. It reads redacted local JSON supplied by the user.
What does an empty steps array mean?
It means normal execution evidence is absent. It does not prove a secret or application code failed.
Can it classify a normal step failure?
Yes. A recorded failing step takes precedence over pre-step inference.
Does it edit workflow YAML?
No. It returns next checks only.
Can it diagnose every platform incident?
No. It labels unsupported or contradictory evidence unresolved.
When should I use CI Doctor?
Use the paid CI Doctor when the next job requires repository, workflow-revision, annotation, matrix, manifest, or related-run inspection.
CI Failure Snapshot keeps proof and approval boundaries visible.
The listing includes the tested package, realistic samples, declared permissions, and known limitations.
Get CI Failure Snapshot on Agensi