wiki_tag_synclisted

# LLM Wiki — Tag & Alias Sync Skill (wiki_tag_sync) > **CLI (read first):** This skill drives the `magi` CLI (MAGI research workspace tool, assumed installed on PATH). If unsure of your surroundings, run `magi sync` first to locate the workspace. For the full syntax of any command: `magi <command> --help`. This skill resolves the "vocabulary fragmentation" problem. Over time, different agents might invent slightly different tags or aliases for the exact same physical/mathematical concept (e.g., `qca`, `quantum-cellular-automata`, `clifford-qca`). This skill acts as a Map-Reduce pipeline to canonicalize them. > **Tooling (framework-agnostic):** This skill is written tool-agnostic. Where it says *file-read tool*, use your agent's equivalent (`Read` in Claude Code, `view_file` in Antigravity). Shell commands run via `Bash`/`PowerShell` (Claude Code) or your framework's shell tool. When the user asks to sync, reduce, deduplicate, or normalize tags and aliases: ### 1. Map Phase (Extraction) * **Action**: Run the deterministic extractor to gather all unique metadata. `magi tags extract <TOPIC_DIR>` * **Result**: This will silently generate two statistical inverted index files: 1. `<TOPIC_DIR>/scratch/raw_tags.json` 2. `<TOPIC_DIR>/scratch/raw_aliases.json` *(These JSONs output an inverted index in the format: `{"Tag_Name": {"count": 2, "files": ["wiki/concepts/file.md"]}}`)* ### 2. Reduce Phase (LLM Decision) * **Action**: Read the contents of both `raw_