Doc Search

BM25 inverted index + LLM query expansion + grep. Zero external dependencies.

Workflow

Follow these steps in order every time:

Step 1 — Resolve docs_dir

Check if the user mentioned a docs directory. If not, ask:

"Which directory should I search? (e.g. ~/obsidian, ~/notes)"

Step 2 — Check index exists

ls \x3Cdocs_dir>/.cache/index.json

• If missing → go to Step 3 (build index first)
• If exists → skip to Step 4

Step 3 — Build index

python3 ~/.claude/skills/doc-search/scripts/build_index.py \x3Cdocs_dir>

Index saved to \x3Cdocs_dir>/.doc-search/index.json. Incremental on subsequent runs.

Step 4 — Expand query terms

Before searching, expand the user's query to cover synonyms, Chinese/English variants, and likely headings. Combine into one string:

"获取音色列表" → "获取 查询 list voice 音色 tts ListVoice 音色列表"

Step 5 — BM25 search

python3 ~/.claude/skills/doc-search/scripts/search.py "\x3Cexpanded query>" \
  --docs-dir \x3Cdocs_dir> --topk 5

Output: JSON array [{path, rel, score, title, summary}, ...]

Step 6 — Grep top-K for precise location

For each result file, grep with the original keywords:

grep -ni -e "keyword1" -e "keyword2" /path/to/doc.md

Report file path + matching line numbers to the user.

Options