agentleFS
inttegrointtegro/labor/skills/production-verification/SKILL.md

production-verification

Skill1 starsChanged 4 days agoMITView on GitHub

Use for production incidents, deployment verification, readiness questions, provider or account-state checks, live-data investigations, migrations, backfills, and claims that a change is deployed, fixed, live, or production-ready. Build evidence across source, tests, revisions, runtime configuration, external state, and observed behavior without turning hypotheses into facts.

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 6 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Verify the exact claim the user cares about.

    line 8
  2. Distinctive:

    Do not skip levels in the report.

    line 29
  3. Distinctive:

    Start production investigations read-only.

    line 33
  4. Distinctive:

    Treat documentation, dashboard labels, support statements, and accepted submissions as evidence about those artifacts, not automatic proof …

    line 35
  5. Distinctive:

    Keep credentials, tokens, private keys, request bodies, authorization headers, signed URLs, and unnecessary customer data out of commands, …

    line 36
  6. Distinctive:

    Use exact timestamps and time zones when correlating distributed systems.

    line 37

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Production Verification

Verify the exact claim the user cares about. Source code, local tests, a commit, a successful push, a deployment record, and confirmed live behavior are different evidence levels.

Choose the relevant procedure

  • For incidents, incorrect production data, or causal diagnosis, read references/investigations.md.
  • For deployment, rollout, readiness, or certification questions, read references/deployment-and-readiness.md.
  • For provider correspondence, capability disputes, or follow-up monitoring, read references/provider-operations.md.

Evidence ladder

Track the highest level actually proven:

  1. proposed or designed;
  2. implemented in the working tree;
  3. focused local validation passed;
  4. committed and history verified;
  5. pushed to the intended remote ref;
  6. built and deployed as an identified revision;
  7. runtime configuration and migrations verified;
  8. live behavior and resulting data reconciled.

Do not skip levels in the report. A lower level can still be useful, but label it accurately.

Safety boundary

  • Start production investigations read-only. Do not mutate records, retry transactions, backfill data, rotate credentials, alter provider settings, or deploy fixes without explicit authorization.
  • Form a concrete causal path, then call it a hypothesis until records, logs, revisions, and external state confirm it.
  • Treat documentation, dashboard labels, support statements, and accepted submissions as evidence about those artifacts, not automatic proof of runtime behavior.
  • Keep credentials, tokens, private keys, request bodies, authorization headers, signed URLs, and unnecessary customer data out of commands, diagnostics, and reports.
  • Use exact timestamps and time zones when correlating distributed systems.

Completion standard

Report confirmed facts, remaining hypotheses, checks that could not run, required authorization, and the highest evidence level reached. Never use done, fixed, live, or production-ready when the evidence supports only implementation or local validation.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.