agentleFS
Sign inStart free
matthewwong525matthewwong525/WongStack/.agents/skills/verify/SKILL.md

verify

Skill7 starsChanged 11 days agoMITView on GitHub

Check previews or CI behavior, report results and remaining help; also tests APIs or screenshots.

  • Reads credentials

188 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use instinct-export, instinct-import 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.
  • TheKREATORSSkill

    instinct-export

    Export instincts from project/global scope to a file

    Also in 3 projects that use this

  • TheKREATORSSkill

    instinct-import

    Import instincts from file or URL into project/global scope

    Also in 3 projects that use this

  • AviadPSkill

    risk-assess

    Assess the risk and blast radius of a proposed code change before implementing it. Evaluates CI impact, framework-wide vs localized scope, rollback path, and identifies blindspots. Use when user says "what's the risk", "is this safe", "assess the impact", "risk assess", "blast radius", or before applying framework-wide or CI-affecting changes. Do NOT use for code quality review (use code-quality) or plan validation (use design-validator).

    Also in 3 projects that use this

  • diego-tobalinaSkill

    build-fix

    Fix build and compilation errors systematically. Run when build fails.

    Also in 6 projects that use this

  • AaravChadhaSkill

    secure

    Security review with a confidence gate - a finding exists only with a concrete exploit scenario and a high/medium/low confidence rating; everything else goes to a worth-hardening list, never inflated. Sweeps five surfaces: auth gates, secrets hygiene, injection and unsafe sinks, unsafe deserialization/crypto/transport, and LLM tool-use trust boundaries. Reports only, never fixes. Use when the user asks for a security review or to check vulnerabilities, secrets, or auth.

    Also in 3 projects that use this

  • ArvorCoSkill

    specify

    Create feature specification from natural language description. Use when starting a new feature or creating a spec. Triggers on: create spec, specify feature, new feature spec.

    Also in 10 projects that use this

  • +1 more from TheKREATORS

The file

We split this skill into the rules it gives an agent

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

  • 0

    Common

    Most of the 167 edited versions we read keep these.

  • 2

    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
  • could be automatic: a setting could enforce it
  1. Distinctive:

    Check fresh scenarios on the exact saved revision.

    line 9
  2. Distinctive:

    No token/surviving block → unverified.

    line 33
  3. Very specific:

    Keep journeys/evidence outside the repo, bar /ship's kept checks.

    line 55
  4. Very specific:

    Never merge/archive: /ship owns that.

    line 56

“Common” is measured: kept by at least 6 in 10 of the 167 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

/verify

Check fresh scenarios on the exact saved revision. Reference: how; wiki: why. Gates nothing.

Authorized: needed /save (kept checks' too), browser install, Access heal, staging preparation and § a's writes; other permissions join the handoff.

Order

ROOT="$(git rev-parse --show-toplevel)"
bash "$ROOT/.claude/skills/verify/scripts/verify-staging.sh" scout-check
bash "$ROOT/.claude/skills/verify/scripts/verify-staging.sh" preflight
  1. Scout the change this chat or the caller named, else the branch's own, archived included (rungs). Asks name candidates' scenarios. READY → match probes/recipes/consumers by § a; nothing reachable → NONE, name exclusions and stop.
  2. Bind. Run node "$ROOT/.claude/skills/save/scripts/saved-revision.mjs" with the caller's --checkpoint <file>. NEEDS_SAVE → /save; SAVED → reuse its matching receipt or read the existing gate. UNKNOWN → unverified. Never save to find out, edit records, or rerun settled checks: each starts another round of checks. Prepare independently. Then preflight (allow 12 minutes for the staging turn); no browser journeys → --no-browser. SEEDED=no beside a TURN line → staging-data checks unverified, with its reason; report TURN=unavailable. CI-only/missing preview → --no-preview --no-browser, then collect. A blocked or failed prerequisite pauses only its dependents; never guess URLs: one can answer for a commit never deployed. Recheck head before posting; changes require fresh evidence.
  3. Walk/grade by §§ b–f; finish independent safe checks before repairs. Inside /ship, add kept checks.
  4. FAILURE → judge scope. In scope → collect failures, fix once, /save, preflight, repeat affected checks at most twice; retain independent evidence only with unchanged dependencies; otherwise report.
  5. Simulate/report by § f; one report covers surfaces/retries, heals/fixes/installs and scope/limits. Ask the next step with the question tool; inside /ship, return the verdict.

When a block stops the walk

Heal once per invocation, never on BLOCK=not-own-site; retry dependents only.

BLOCK=access-challenge (exit 3): mint a repo-named service token with the Cloudflare API token, widening if needed. Confirm the preview app's machine policy accepts it; keep human permissions. Store it as CF_ACCESS_STAGING_CLIENT_ID/SECRET in the primary worktree's .env only (secrets, Access, authorized widen). Name the mint; never print/commit credentials. Privacy changes need human login/revocation checks; machine access proves no email login.

No token/surviving block → unverified; name credentials/repair gaps.

Plain checks

Named screenshot/request/click-through: skip scout, use its address and mktemp -d "${TMPDIR:-/tmp}/wong-verify-XXXXXX"; otherwise bind/preflight. Same safeguards/handoff. Past pictures: verify-staging.sh pictures <pr>. Show evidence; post only if asked; cleanup.

Verdicts

VerdictMeaningReport
NONEno scenario reachablewhat was there instead
SUCCESSno contradiction or blocked reachable checkevidence, including each partly shown claim
FAILUREevidence contradicts a THENevidence, fix only in scope
UNKNOWNreachable check blocked or ambiguousunverified, why and needed help
TIMEOUTbudget exceededunverified, what completed and where it stopped

Precedence: FAILURE, TIMEOUT, UNKNOWN. UNKNOWN is not NONE (why).

Hard rules

  • agent-browser/Chrome only for browser journeys. New probe runtimes ask first; bundled collector authorized. No repo dependencies.
  • Saved-source previews/validated CI captures; no local repo execution/invented tooling. Read-only recipes; written expectations judge.
  • Keep journeys/evidence outside the repo, bar /ship's kept checks; cleanup on every exit, including UNKNOWN/handoff.
  • Never merge/archive: /ship owns that.

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