← ClaudeAtlas

plugin-validatelisted

Validates a generated plugin folder against validator.config.json rules — line limits, naming conventions, and skill-agent pairing.
EngAhmedShehatah/plugin-architect · ★ 1 · AI & Automation · score 70
Install: claude install-skill EngAhmedShehatah/plugin-architect
## What this skill does Reads every artifact inside `plugin_path` and checks it against the rules defined in `validator.config.json`. Returns a structured report of all violations. ## How to execute this skill This skill is fully self-contained and works standalone on any tool. Execute each step in order using whatever methods are available to you (file reading, etc.). 1. Load the validator config from `config_path` (or use the default path if not provided) 2. Recursively read all `.md` files under `plugin_path/skills/` and `plugin_path/agents/` 3. For each file, apply all 8 validation rules (detailed below) 4. Collect violations and their severity levels 5. Generate a structured violation report 6. Return a single JSON object only — do not append prose or a separate text summary **Error handling:** - If `plugin_path` does not exist, report this as a critical error and stop - If `validator.config.json` cannot be read, use the default rules built into the skill You can run this skill entirely on your own — no agent orchestration is required. **Output formatting:** - `passed: true` only when there are zero `error`-severity violations - Return violations array sorted by severity (errors first) then by file path - Keep the `summary` field inside the JSON object as the only human-readable summary - Do not append any prose after the JSON object ## Validation rules ### Rule 1 — Line limit (`max-lines`) Every `.md` file inside `plugin_path` must not exceed the `max_lines`