# build-skill

build-skill walks a new Claude skill through four agents: requirement questions, file generation, validation, docs. You approve the spec before any file appears.

- command · `type=command` · Claude Code, Codex, GitHub Copilot, Gemini CLI · ★ 1357 repo stars
- Author: qdhenry
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/qdhenry/Claude-Command-Suite/tree/HEAD/.claude/commands/skills/build-skill.md
- Updated: 2026-03-01
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/commands/build-skill

## What it does

The `/build-skill` command chains four subagents through the Task tool. `skill-elicitation-agent` asks 3-5 questions about purpose, scope, tool permissions and success criteria, then turns the answers into a spec. Nothing gets generated until you sign off on it.

\`skill-generator-agent` then lays out the folder: `SKILL.md` with frontmatter, plus `reference.md`, `examples.md`, `scripts/` and `templates/` where the spec asks for them. The target can be a personal skill under `~/.claude/skills/`, a project skill or a plugin skill.

\`skill-validator-agent` reads the YAML, checks whether the description carries the keywords that make a skill trigger, follows file references and tests script syntax. It scores each area and sends failures back to the generator for another pass. `skill-documenter-agent` closes with a README, worked examples and troubleshooting notes. The whole chain assumes those four agents exist in your setup, because the command calls them by name through Task; with none of them installed there is nothing to launch.

## Install

### Claude Code (native platform)

**Manually: put the files in place**
- File: `.claude/commands/build-skill.md` — Call it with /build-skill. For all projects use ~/.claude/commands/

### Codex

**Manually: put the files in place**
- File: `~/.codex/prompts/build-skill.md` — Call it with /prompts:build-skill. Codex marks this format deprecated in favour of skills

### GitHub Copilot

**Manually: put the files in place**
- File: `.github/prompts/build-skill.prompt.md`

### Gemini CLI

**Manually: put the files in place**
- File: `.gemini/commands/build-skill.toml` — Gemini stores commands as TOML: the prompt text goes into the prompt field

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/qdhenry/Claude-Command-Suite/tree/HEAD/.claude/commands/skills/build-skill.md

## When to use it
- scaffold a `SKILL.md` whose frontmatter parses
- turn a rough idea into a written skill spec
- check trigger keywords before a skill goes live
- lay out scripts and templates folders correctly
- add a README and examples to a finished skill

## Example prompts
- \`/build-skill` I want a skill that checks Alembic migrations before each commit
- \`/build-skill` build a changelog skill from git log, project level
- \`/build-skill` make a skill with a Python script that validates JSON schemas, personal folder

## Similar
- [recoup-platform-build-os](https://affpapa.org/en/nevibecode/skills/recoup-platform-build-os.md) — skill · 5200 installs
- [build](https://affpapa.org/en/nevibecode/commands/disler-build.md) — command · ★ 3990 repo stars
- [build-zoom-apps-sdk-app](https://affpapa.org/en/nevibecode/commands/build-zoom-apps-sdk-app.md) — command · ★ 8052 repo stars
- [google-agents-cli-scaffold](https://affpapa.org/en/nevibecode/skills/google-agents-cli-scaffold.md) — skill · 585900 installs
- [convex-quickstart](https://affpapa.org/en/nevibecode/skills/convex-quickstart.md) — skill · 139100 installs
- [init-project](https://affpapa.org/en/nevibecode/commands/init-project.md) — command · 1382 installs

From the same set: [qdhenry/claude-command-suite](https://affpapa.org/en/nevibecode/by/qdhenry/claude-command-suite.md) — 112 cards

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
