agentleFS
Sign inStart free
maidepsmaideps/ai-harness-framework/skills/verification/SKILL.md

verification

Skill2 starsChanged 6 days agoMITView on GitHub

Running the verification gates honestly and reading their output, not just exit codes.

47 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use evidence-driven-writing, experiment-results-planning and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • Norman-burySkill

    evidence-driven-writing

    Use when writing or revising Introduction, Related Work, background, literature synthesis, or any section where references must drive claims

    Also in 3 projects that use this

  • Norman-burySkill

    experiment-results-planning

    Use when designing experiments, result tables, mock planning data, evaluation protocols, or results sections before real data are final

    Also in 3 projects that use this

  • +10 more from Norman-bury

The file

We split this skill into the rules it gives an agent

7 rules. 0 are common, 5 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 39 edited versions we read keep these.

  • 5

    Distinctive

    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. Distinctive:

    Before claiming any work is done

    line 12
  2. Distinctive:

    When a layer fails and you need to know what actually broke

    line 13
  3. Very specific:

    Run the full chain in order: npm run check (lint → typecheck → test → build → e2e → verify-all → quality).

    line 17
  4. Distinctive:

    Read the output, not only the exit status:

    line 18
  5. Very specific:

    Before committing: npm run clean-check (harness files present, working tree clean, no debug artifacts, WIP ≤ 1, no OS artifacts staged).

    line 23
  6. Distinctive:

    Never claim done without running the required commands.

    line 29
  7. Distinctive:

    Do not skip layers to save time.

    line 31

“Common” is measured: kept by at least 6 in 10 of the 39 edited versions of this file we could read. “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

How to run the gates and what the tri-state results mean.

When to Use

  • Before every commit
  • Before claiming any work is done
  • When a layer fails and you need to know what actually broke

Workflow

  1. Run the full chain in order: npm run check (lint → typecheck → test → build → e2e → verify-all → quality).
  2. Read the output, not only the exit status:
    • PASS — the layer ran and succeeded; it counts as verified.
    • SKIP — the layer is not configured; it does NOT count as verified.
    • FAIL — non-zero exit; the chain stops. Fix and rerun the failing layer, then the chain from that layer on.
  3. For architecture enforcement: npm run check-arch (required surfaces, WIP=1/dependency order, no placeholder docs, valid manifest, every tracked file classified).
  4. Before committing: npm run clean-check (harness files present, working tree clean, no debug artifacts, WIP ≤ 1, no OS artifacts staged).
  5. If the project enabled quality gates (.harness/quality.json has "enabled": true), npm run quality runs last in the chain. Read each gate line: [PASS] coverage: 84.20 ≥ 80 counts; [SKIP] mutation: disabled does not; a [FAIL] names the metric and the floor it missed. Never lower a min to make a gate pass without saying so to a human.
  6. For a specific feature: npm run verify-feature -- <feature-id> — runs the feature's own layers in order, stops at the first failure, prints the repair hint, and records evidence only when all layers pass. It refuses while code is uncommitted (harness bookkeeping excepted) and fails if a layer leaves files behind, so the commit in the evidence is the code that passed. A [WARN] … layers changed after it was verified line from npm run typecheck means that feature's evidence is stale: re-run its gate.

Rules

  • Never claim done without running the required commands.
  • A SKIP is honest — it is not a pass. Do not present a SKIP as verification.
  • Do not skip layers to save time; each layer gates the next.

Quick Reference

  • npm run check — full chain, output read
  • npm run quality — metric floors, when enabled
  • npm run check-arch — architecture rules
  • npm run clean-check — pre-commit clean state
  • PASS ≠ SKIP; report both honestly

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.

Check your setup, free