agentleFS
nuncaeslupusnuncaeslupus/literary-critic/.claude/skills/gate-check/SKILL.md

gate-check

Skill0 starsChanged 6 days agoMITView on GitHub

Reports PASS/FAIL with measured numbers on a plan task's acceptance gate, or audits all gates. Use when the user asks whether a gate passed. Not for writing code.

22 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use project-stage-detect, setup-engine 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.
  • CoralGameSkill

    project-stage-detect

    Automatically analyze project state, detect stage, identify gaps, and recommend next steps based on existing artifacts.

    Also in 17 projects that use this

  • CoralGameSkill

    setup-engine

    Configure the project's game engine and version. Pins the engine in CLAUDE.md, detects knowledge gaps, and populates engine reference docs via WebSearch when the version is beyond the LLM's training data.

    Also in 13 projects that use this

  • lilangMaxSkill

    asset-spec

    Generate per-asset visual specifications and AI generation prompts from GDDs, level docs, or character profiles. Produces structured spec files and updates the master asset manifest. Run after art bible and GDD/level design are approved, before production begins.

    Also in 10 projects that use this

  • lilangMaxSkill

    create-control-manifest

    After architecture is complete, produces a flat actionable rules sheet for programmers — what you must do, what you must never do, per system and per layer. Extracted from all Accepted ADRs, technical preferences, and engine reference docs. More immediately actionable than ADRs (which explain why).

    Also in 10 projects that use this

  • +3 more from CoralGame
  • +5 more from lilangMax

The file

We split this skill into the rules it gives an agent

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

  • 0

    Common

    Most of the 21 edited versions we read keep these.

  • 0

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

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

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

    Read a task's measurable acceptance gate from status/plan.md, compare a measured value to its threshold, and report PASS/FAIL with the …

    line 11

“Common” is measured: kept by at least 6 in 10 of the 21 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

gate-check

Read a task's measurable acceptance gate from status/plan.md, compare a measured value to its threshold, and report PASS/FAIL with the numbers.

CANARY: gate-check-loaded-2026-06-06-fb78d23e-e26f5d00ce664711

When to load

Load when a plan already carries a Gate column and the question is whether a finished task met it, with the measured number. execution writes the evidence; this skill verifies it, and review / ship run it.

The gate contract

A gate is a measurable acceptance condition in the plan's task table, written <metric> <op> <threshold> with one of < <= > >= == !=:

| T# | Description | Gate                    | Tests |
|----|-------------|-------------------------|-------|
| T1 | parser      | `line_coverage >= 0.90` | …     |
| T2 | endpoint    | `p95_latency_ms <= 200` | …     |

The evidence (measured value, command, commit SHA, environment provenance) lives in the plan's Evidence log table. Load references/gate-grammar.md when writing a gate the parser must read, defining the Evidence log columns, or diagnosing a gate reported "manual" or a task reported "grandfathered".

How to use

The script defaults to status/plan.md; pass --input for any other path.

Audit every task

python3 "${CLAUDE_SKILL_DIR}/scripts/run_gate.py" --input status/plan.md

Prints one line per task and a summary. Exit 0 when no gated task fails or lacks evidence, 1 otherwise. A manual (?) gate exits 0 once its evidence is recorded, so it still needs a human verdict before it counts as met. --strict also fails ungated tasks; without it they are grandfathered.

Focus one task, compare a measured value

python3 "${CLAUDE_SKILL_DIR}/scripts/run_gate.py" --input status/plan.md --id T2 187

Compares the measured 187 to T2's threshold; omit the number to use the value recorded in the Evidence log. --json gives a verdict a wrapper can consume: a project's own wrapper measures the metric, then calls run_gate.py for the verdict.

run_gate.py is read-only, so checking a gate never changes the record it audits; recording evidence is a markdown edit owned by execution's RECORD step.

Gotchas

  • Threshold and measured value must share a unit. The parser reads the first number after the operator (>= 90% → 90); it does not convert. A gate of >= 90 checked against a measured 0.93 fails — pick one scale and keep the measurement on it.
  • Non-numeric gates report "manual", never PASS. A gate the grammar cannot parse (e.g. "all golden files identical") is surfaced for human judgement and is never auto-passed — rewrite it as a measurable condition where one exists.
  • Exit 2 means no Gate column or a usage error (missing file, bad --id). A wrong --input gives the same code, so confirm the path before treating it as a plan that predates gates.
  • Stale evidence after a re-run. The Evidence log records one measurement at a point in time (its SHA + provenance say which). After re-measuring, update the row — run_gate.py trusts the recorded number; it does not re-run the command.

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.