agentleFS
Cody-SimsCody-Sims/agent-skills/skills/verification-discipline/SKILL.md

verification-discipline

Skill0 starsChanged 4 months agoMITView on GitHub

Audits and hardens existing tests and CI gates against vacuous assertions, copied production logic, unbounded checks, unenforced invariants, and dirty builds. Use when investigating suspicious green tests, repairing test credibility, or enforcing validation; not for ordinary test-first features or routine completion checks.

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Bound the proof: Read applicable testing instructions.

    line 18
  2. Very specific:

    Prove sensitivity: Use a reversible negative control matching the target: mutate production for behavior assertions, or isolated …

    line 26
  3. Distinctive:

    Verify and stop: Run the focused proof, then the relevant full gate on final state, not repeated unchanged gates.

    line 40

≈ “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

Verification Discipline

A green test must catch its claimed bug. Never weaken assertions, broaden mocks, skip cases, or copy production logic to obtain green; failures are findings.

Workflow

  1. Bound the proof. Read applicable testing instructions; name the protected behavior/invariant/build property. Run the narrow existing baseline and inspect its production call/assertion. Stay within this target; do not chase coverage.
  2. Strengthen observation. Assert externally visible behavior, fixed examples, independent fixtures, or invariants through the real module, not just names, return types, or absence of a throw. Read red-green proof when adding/strengthening assertions or investigating vacuity.
  3. Prove sensitivity. Use a reversible negative control matching the target: mutate production for behavior assertions, or isolated CI/config/build inputs for gate and build properties. The real guard must fail on the intended assertion, then pass after restoring only that control. Never overwrite unrelated work or commit the mutation. If unsafe, mark proof blocked. Still-green mutants need stronger tests; real bugs need fixes or blockers, never looser expectations.
  4. Enforce relevant gates. Required gates must invoke durable invariant tests, not leave them in unused scripts. Bound CPU workloads and set explicit per-test timeouts. Compare worktree before/after builds and inspect tracked churn: a defect or explicit owning-gate decision, not consent to reset unrelated work. Read invariant and gate patterns for CI enforcement, timeouts, report-only adoption, or build purity.
  5. Verify and stop. Run the focused proof, then the relevant full gate on final state, not repeated unchanged gates. Return behavior, commands/results, mutation/red-green evidence, exposed bugs, timeout/report-only changes, and build cleanliness relative to baseline. Mark unrun/failed checks blocked; green alone is not credibility.

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.