Skip to content
Back to skills

Astro Skill Lint

ASecurity

Validates ASTROLABE project skills fail-closed — frontmatter parses, name matches directory, description within budget (≤1024 chars, ≤1536 combined with when_to_use), body ≤500 lines, referenced files exist, no unknown frontmatter keys, no command/skill name collisions. Use after creating or editing anything under .claude/skills or .claude/commands, and before committing skill changes.

  • 4 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added September 22, 2026
ai-agentspythonbashtesting

Works with

  • claude code

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned September 22, 2026

npx -y skills add ChrisRoyse/Astrolabe --skill astro-skill-lint --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Astro Skill Lint?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Astro Skill Lint
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/chrisroyse-astro-skill-lint/badge)](https://www.skillsdirectory.com/skills/chrisroyse-astro-skill-lint)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: astro-skill-lint
description: Validates ASTROLABE project skills fail-closed — frontmatter parses, name matches directory, description within budget (≤1024 chars, ≤1536 combined with when_to_use), body ≤500 lines, referenced files exist, no unknown frontmatter keys, no command/skill name collisions. Use after creating or editing anything under .claude/skills or .claude/commands, and before committing skill changes.
allowed-tools: Bash(python *)
---

# Skill linter

Run from the repo root:

```
python ${CLAUDE_PROJECT_DIR}/.claude/skills/astro-skill-lint/scripts/lint_skills.py
```

- Exit 0 = every skill valid (per-skill `OK` lines + summary).
- Exit 1 = violations, one `FAIL <path>: <message>` line each. Fix and re-run until green; only proceed (commit, close) when the lint passes.
- An alternate skills root can be passed as the first argument (used for fixture testing the linter itself).

## Standing rules

- Every edit under `.claude/skills/` or `.claude/commands/` ends with a green lint run; paste the summary line in the commit-referenced issue comment.
- Malformed YAML fails soft at runtime (Claude Code loads the body with empty metadata and auto-invocation silently dies) — this linter is the only loud guard; treat its failure as a broken build.
- Keep bodies lean: content loaded once persists in context all session. Detail belongs in `references/` (one level deep), deterministic logic in `scripts/` with verbose fail-closed errors.
- New shared conventions (telemetry kinds, gate codes) update the owning reference file in the same commit.

Files in this skill

  • SKILL.md1.6 KB
  • scripts/lint_skills.py6.2 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.