gan SKILL.md skillA

gan from gufranco/claude-engineering-rules is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned
mdr add gufranco/claude-engineering-rules/gan@git:20261012.84ebb18

Pins git:20261012.84ebb18 in mdr.lock. Pin the fingerprint instead.

Audit grade
A, 17 of 17 checks passed
Kind
SKILL.md skill
Repository
gufranco/claude-engineering-rules
Path
skills/gan/SKILL.md
Versions recorded
3
Latest version
git:20261012.84ebb18, committed Oct 12, 2026
First version recorded
May 25, 2026
Size
6,832 bytes
Repository stars
3
License
MIT

Generator-evaluator loop that iterates a feature build until it clears a scored rubric threshold.

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

## Overview

Run a planner → generator → evaluator loop with a score-gated stop condition. The planner expands the brief into acceptance criteria and a rubric. The generator implements against the criteria. The evaluator scores the result, returns feedback. The loop iterates until the score crosses the threshold or the iteration cap is hit.

## When to use

- The output has a measurable quality signal, Playwright test outcome, lint pass, custom rubric.
- The first attempt is unlikely to be the last attempt.
- The cost of running a feedback loop is justified by the quality lift.

## When NOT to use

- One-shot implementations where the spec is unambiguous.
- Recurring schedules. Use the `/loop` built-in.
- Visual design exploration. Use `/design variants`.
- Trivial fixes. Direct edit is faster.

## Arguments

`/gan "<brief>"` is the canonical form. The brief is one or two sentences describing what to build.

Environment variables:

- `GAN_MAX_ITERATIONS` (default 8). Hard cap on planner-generator-evaluator cycles.
- `GAN_PASS_THRESHOLD` (default 8.0). Weighted score 0-10 to consider passing.
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint
mdr add gufranco/claude-engineering-rules/gan@sha256:afe4fc9f4b8aeb68

The line at the top of this page pins the label git:20261012.84ebb18, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_5km7rnnpcah45p2k.svg)](https://markdownregistry.com/a/art_5km7rnnpcah45p2k)

3 badge image fetches in 30 days. What the badge shows.

Versions

The registry has recorded 3 versions of this file, between May 25, 2026 and Oct 12, 2026, about one every 70 days.

versioncommittedcommitsizeaudit
git:20261012.84ebb18 latest2026-10-12 84ebb18 6,832 BA view · diff
git:20260527.6d7d5622026-05-27 6d7d562 5,469 BA view · diff
git:20260525.41f34932026-05-25 41f3493 5,683 BA view

Audit of the latest version

A  17 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (6832 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

gufranco/claude-engineering-rules · 3 stars · license MIT · pushed 2026-10-30 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_5km7rnnpcah45p2k
GET https://markdownregistry.com/api/v1/resolve?ref=gufranco/claude-engineering-rules/gan
GET https://markdownregistry.com/api/v1/blob/afe4fc9f4b8aeb6806faa23057649ab6b62bcb78f527e1452ce15adbb5f2519f

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.

More from gufranco/claude-engineering-rules

CLAUDE.md CLAUDE.md file
gufranco/claude-engineering-rules · CLAUDE.md
git:20261029.8b030ee · audit A · 3 stars
aside SKILL.md skill
gufranco/claude-engineering-rules · skills/aside/SKILL.md · Mid-task side question handler. Freezes the active task state, answers a side question without state mutation, then…
git:20261012.84ebb18 · audit A · 3 stars
assessment SKILL.md skill
gufranco/claude-engineering-rules · skills/assessment/SKILL.md · Audit an implementation's architecture for completeness, missing patterns, and planted defects before a take-home or…
git:20261029.efb0280 · audit A · 3 stars
audit SKILL.md skill
gufranco/claude-engineering-rules · skills/audit/SKILL.md · Security audit across dependencies, secrets, Docker, code patterns, images, STRIDE, and OWASP. Covers dependency…
git:20261029.9c8290b · audit B · 3 stars
benchmark SKILL.md skill
gufranco/claude-engineering-rules · skills/benchmark/SKILL.md · Measure API latency, bundle size, and database query time, and compare them against saved baselines to catch…
git:20261013.02cb8cd · audit A · 3 stars
brain SKILL.md skill
gufranco/claude-engineering-rules · skills/brain/SKILL.md · Read from and write to the second brain vault. Subcommands - capture (default), ingest, ask, link, health, refresh…
git:20261022.be3b951 · audit A · 3 stars
checkpoint SKILL.md skill
gufranco/claude-engineering-rules · skills/checkpoint/SKILL.md · Save and resume working state across sessions, with subcommands save, resume, and list. Captures git state, decisions…
git:20261013.02cb8cd · audit A · 3 stars
cleanup SKILL.md skill
gufranco/claude-engineering-rules · skills/cleanup/SKILL.md · List stale branches, long-open PRs, merged branches, and orphaned worktrees, then close or delete each one with…
git:20261013.02cb8cd · audit A · 3 stars
concise-mode SKILL.md skill
gufranco/claude-engineering-rules · skills/concise-mode/SKILL.md · Opt-in terse-reply mode that drops articles, pleasantries, filler, and hedging while keeping full technical accuracy…
git:20261012.84ebb18 · audit A · 3 stars
cross-model SKILL.md skill
gufranco/claude-engineering-rules · skills/cross-model/SKILL.md · Consult a different model for orthogonal reasoning on a consequential decision, especially when confident. Discovers…
git:20261012.84ebb18 · audit A · 3 stars
deploy SKILL.md skill
gufranco/claude-engineering-rules · skills/deploy/SKILL.md · Merge a shipped PR, verify the deployment, and monitor post-deploy health, with land and canary subcommands.
git:20261013.02cb8cd · audit A · 3 stars
design SKILL.md skill
gufranco/claude-engineering-rules · skills/design/SKILL.md · Design consultation, variant exploration, and design system scaffolding, with subcommands consult (default), variants…
git:20261013.02cb8cd · audit A · 3 stars

Every file in gufranco/claude-engineering-rules

Browse by kind, by grade A, or by owner.