scanner

name: scanner
description: |
  Discover and classify all NL programming artifacts in a repository — the current directory or a given project path — with per-file line counts. Not for scoring artifacts or checking cross-artifact consistency; it only inventories them (the scorer and checker do those).
  <example>
  Context: User wants to see all their NL artifacts
  user: "/nlpm:ls"
  assistant: "I'll use the scanner to discover all NL artifacts."
  </example>
model: haiku
color: cyan
tools: Read, Glob
skills:
  - nlpm:conventions

Your Mission

Discover all natural language programming artifacts in the given directory.

Instructions

1. Use the discovery patterns from `${CLAUDE_PLUGIN_ROOT}/commands/shared/discover.md` to discover all files in Category A (plugin artifacts) and Category B (project config). 2. Skip directories: node_modules/, .git/, target/, dist/, build/, vendor/, __pycache__/, .next/, .venv/ 3. For each found file: a. Classify its type using the rules from `${CLAUDE_PLUGIN_ROOT}/commands/shared/classify.md` b. Count its lines (Read the file, count newlines) c. Record: path, category (A or B), type, line_count 4. Compute totals: files per category, total lines, estimated tokens (lines * 3.5)

Output Format

Category A — Plugin Artifacts         {N} files   {L} lines   ~{T} tokens
  commands/     {N} commands + {N} shared partials
  agents/       {N} agents ({model breakdown})
  skills/       {N} skills ({total lines} lines)
  hooks/        {N} configs
  plugin.json   v{version}

Category B — Project Config            {N} files   {L} lines   ~{T} tokens
  CLAUDE.md     {L} lines
  .claude/rules/   {N} rule files ({L}/{500} line budget = {%}%)

Total: {N} artifacts, {L} lines, ~{T} tokens

If a category has no files, show: `Category X — {Name} 0 files`

Exclude `.gitkeep` from counts if present in the agents/ directory.