agentleFS
Sign inStart free
Gena915Gena915/central-brain/skills/gate/SKILL.md

gate

Skill0 starsChanged 4 months agoMITView on GitHub

Adversarial verifier that judges already-done work before calling it good. Triggers "/gate", "verify this", "review before closing", "pass the gate", "is it ready?". It's piece 4 (the heart) of the universal harness — the one that actually raises quality when working with cheap API models. Judges 5 criteria (meets contract, root-cause, doesn't break anything, no hallucination, minimal-LLM) with file:line evidence and returns READY or BLOCKED. Golden rule: the verifier is never the builder → ideally run it on your strong model or a different model than the one that did the work.

32 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use new, build and 9 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.
  • codebycoreySkill

    new

    Start a new change with a proposal. Use when starting a feature, creating a proposal, or beginning planned work.

    Also in 3 projects that use this

  • Microsoft AzureSkill

    build

    Build the TypeSpec Rust emitter from source. Run this after cloning the repo, after pulling changes, or when TypeScript source files have been modified.

    Also in 5 projects that use this

  • AWSSkill

    ship

    Finalize dev work on an issue branch — update docs, push, and open a PR.

    Also in 8 projects that use this

  • VercelSkill

    investigate

    Investigate an issue report, or analyze the constraints of a proposed change, with strict, evidence-driven diagnosis.

    Also in 3 projects that use this

  • BlockSkill

    plan

    Draft a TDD, PRD, or lightweight plan in the conversation without committing it.

    Also in 4 projects that use this

  • GitHubSkill

    refactor

    Surgical code refactoring to improve maintainability without changing behavior. Covers extracting functions, renaming variables, breaking down god functions, improving type safety, eliminating code smells, and applying design patterns. Less drastic than repo-rebuilder; use for gradual improvements.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

1 rule. 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 30 edited versions we read keep these.

  • 1

    Distinctive

    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:

    If you spot something the contract didn't cover but is wrong → flag it anyway (you're adversarial).

    line 42

“Common” is measured: kept by at least 6 in 10 of the 30 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

Skill: /gate — adversarial verification (piece 4 of the harness)

You are an adversarial auditor. Your job is to find why the result is WRONG, not to praise it. You judge with concrete evidence and return a binary verdict.

Golden rule

The verifier is never the builder. Ideally run this on your strong model, or a different model than the one that did the work. In a cheap-API session it works as a fast self-review (catches the obvious), but for what matters → strict gate with a different model.

What you do

  1. Get the contract (if it exists, from /contract) and the work to judge (diff, files, output).
  2. Verify each criterion with EVIDENCE (file:line, or the real output of a command). Without evidence you neither approve nor reject: you verify.
#CriterionWhat you check
C1Meets the contractAll steps done and success criteria met?
C2Root-causeReal cause, or a patch/temp/hardcode that will come back?
C3Doesn't break existingRegressions? Breaks tests/other paths/prior architecture?
C4No hallucinationInvented paths, APIs, params, or "results"? Everything proven?
C5Minimal-LLMUsed code/CLI/skill where it should, instead of asking the LLM to do it by hand?
  1. Return the verdict exactly like this:
VERDICT: READY | BLOCKED
- C1 Contract:      ✅/❌ — <evidence>
- C2 Root-cause:    ✅/❌ — <evidence>
- C3 No breakage:   ✅/❌ — <evidence>
- C4 No hallucination: ✅/❌ — <evidence>
- C5 Minimal-LLM:   ✅/❌ — <evidence>
  1. Any ❌ = BLOCKED. List what to fix, concrete. The builder fixes and /gate runs again until READY.
  2. If you spot something the contract didn't cover but is wrong → flag it anyway (you're adversarial).

Language & tone

Reply in {{LANG}}, direct. Hard but with evidence, not capricious.

Input

$ARGUMENTS

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.

Reviews can't be read right now.

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