agentsclimarketplace

Validate all

Skill pdugan20/claudelint/skills/validate-all

Runs comprehensive claudelint validation on all Claude Code project files. Use when user asks to "check everything", "run all validators", "full audit", "validate my entire project", or "what's wrong with my config". Validates CLAUDE.md, skills, settings, hooks, MCP servers, and plugin manifests.From its SKILL.md

Install
npx -y skills add pdugan20/claudelint --skill validate-all

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 14 stars14 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

3.8 KB, as published. Nobody here has run it

Validate Claude Code Project

Runs claudelint check-all to validate all Claude Code project files including CLAUDE.md, skills, settings, hooks, MCP servers, and plugin manifests.

Usage

claudelint check-all $ARGUMENTS

Options

Available flags:

  • --verbose - Show detailed validation output
  • --warnings-as-errors - Treat warnings as errors
  • --explain - Show detailed explanations and fix suggestions
  • --format <format> - Output format: stylish, json, compact (default: stylish)
  • --config <path> - Path to custom config file
  • --fast - Fast mode: skip expensive checks
  • --color / --no-color - Control color output

Examples

Example 1: Plugin won't load after changes

User says: "I just updated my CLAUDE.md and added a new skill, but now my plugin isn't loading" What happens:

  1. Skill runs claudelint check-all
  2. Finds CLAUDE.md exceeds 50KB (blocking issue)
  3. Finds new skill has wrong capitalization in allowed-tools
  4. Shows both issues need fixing before plugin will load

Result: User identifies why plugin broke and fixes both issues

Example 2: Pre-commit validation failed

User says: "My git pre-commit hook is failing but I don't know which validator is the problem" What happens:

  1. Skill runs claudelint check-all --verbose
  2. Shows validate-mcp failed: invalid transport type "sse" (deprecated)
  3. Shows validate-hooks passed, validate-skills passed, etc.
  4. Points to specific .mcp.json file and line number

Result: User fixes the one failing validator (MCP config) and commit succeeds

Example 3: After npm install, verify everything still works

User says: "I just upgraded claude-code-lint, make sure I didn't break anything" What happens:

  1. Skill runs claudelint check-all
  2. Detects new rules flagging previously-allowed patterns
  3. Shows deprecation warnings for old syntax
  4. Suggests migration path for deprecated features

Result: User knows what needs updating after version upgrade

Command Examples

Basic validation:

claudelint check-all

Verbose output with explanations:

claudelint check-all --verbose --explain

JSON output for CI/CD:

claudelint check-all --format json

Use custom config:

claudelint check-all --config .claudelintrc.custom.json

Treat warnings as errors (strict mode):

claudelint check-all --warnings-as-errors

What Gets Validated

The validate skill checks:

  1. CLAUDE.md files - File size, imports, frontmatter, sections
  2. Skills - SKILL.md schema, naming, documentation, security
  3. Settings - settings.json schema, permissions, environment variables
  4. Hooks - hooks.json schema, events, commands
  5. MCP Servers - .mcp.json schema, transport config, variables
  6. Plugins - plugin.json schema, semantic versioning, references

Exit Codes

  • 0 - No errors or warnings
  • 1 - Warnings found (or warnings treated as errors)
  • 2 - Errors found or fatal error (invalid config, crash)

For troubleshooting, see common issues. For customization, see configuration.

See Also

Keep looking

Skills are one crate of 611,178. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.