agent-sh agnix
Lint agent configurations before they break your workflow. Validates Skills, Hooks, MCP, Memory, Plugins across 10+ AI tools including Claude Code, Cursor, GitHub Copilot, Codex CLI, OpenCode, Gemini CLI, Cline, Windsurf, Kiro, and Amp.
Parse Arguments
const args = '$ARGUMENTS'.split(' ').filter(Boolean);
const targetPath = args.find(a => !a.startsWith('--')) || '.';
const fix = args.includes('--fix');
const strict = args.includes('--strict');
let target = args.find(a => a.startsWith('--target='))?.split('=')[1];
if (!target) {
const idx = args.indexOf('--target');
if (idx !== -1 && args[idx + 1]) target = args[idx + 1];
}
target = target || 'claude-code';
When to Use
Invoke when user asks to:
- "Lint my agent configs"
- "Validate my skills"
- "Check my CLAUDE.md"
- "Validate hooks"
- "Lint MCP configs"
- "Fix agent configuration issues"
- "Check if my SKILL.md is correct"
Prerequisites
agnix must be installed. Check with:
agnix --version
If not installed:
cargo install agnix-cli
Execution
1. Validate Project
agnix .
2. If Issues Found and Fix Requested
agnix --fix .
3. Re-validate to Confirm
agnix .
CLI Reference
| Command | Description |
|---|---|
agnix . |
Validate current project |
agnix --fix . |
Auto-fix issues |
agnix --strict . |
Treat warnings as errors |
agnix --target claude-code . |
Only Claude Code rules |
agnix --target cursor . |
Only Cursor rules |
agnix --watch . |
Watch mode - re-validate on changes |
agnix --format json . |
JSON output |
agnix --format sarif . |
SARIF for GitHub Code Scanning |
Supported Files
| File Type | Examples |
|---|---|
| Skills | SKILL.md |
| Memory | CLAUDE.md, AGENTS.md |
| Hooks | ${STATE_DIR}/settings.json |
| MCP | *.mcp.json |
| Cursor | .cursor/rules/*.mdc |
| Copilot | .github/copilot-instructions.md |
Output Format
CLAUDE.md:15:1 warning: Generic instruction 'Be helpful' [fixable]
help: Remove generic instructions. Claude already knows this.
skills/review/SKILL.md:3:1 error: Invalid name [fixable]
help: Use lowercase letters and hyphens only
Found 1 error, 1 warning (2 fixable)
Exit codes:
0- No errors (warnings allowed)1- Errors found2- Invalid arguments
Rule Categories
| Prefix | Category | Examples |
|---|---|---|
| AS-* | Agent Skills | Name format, triggers, description |
| CC-* | Claude Code | Hooks, memory, plugins |
| MCP-* | MCP Protocol | Server config, tool definitions |
| PE-* | Prompt Engineering | Generic instructions, redundancy |
| XP-* | Cross-Platform | Compatibility across tools |
| AGM-* | AGENTS.md | Structure, sections |
| COP-* | GitHub Copilot | Instructions format |
| CUR-* | Cursor | MDC format, rules |
Common Issues & Fixes
| Issue | Solution |
|---|---|
| Invalid skill name | Use lowercase with hyphens: my-skill |
| Directory/name mismatch | Rename directory to match name: field |
| Generic instructions | Remove "be helpful", "be accurate" |
| Missing trigger phrase | Add "Use when..." to description |
Integration
This skill is standalone and can be invoked directly via /agnix.
For CI integration, see the GitHub Action.