passage-glossary

# Passage Glossary ## Purpose Produce a passage reader artifact: the passage text followed by a deduplicated, MCP-grounded glossary of every distinct headword occurring in it. Data-grounded. Never fabricate glosses. Output to file (default) or inline (`--output print`). **Key constraint:** Every gloss must trace to MCP data (`query_lexicon` or morphology fallback). Unresolved headwords are shown explicitly — never invented. --- ## Headword Identity > **headword identity = strongs ?? lemma** A headword is identified by its **Strong's number** when the morphology data supplies one. When `strongs` is null, the exact `lemma` string serves as the key. This identity rule governs deduplication, lexicon calls, and glossary rendering throughout. --- ## Pipeline (Five Steps — Execute in Order) ### Step 1: Parse References Accept one or more passage references. Supported forms: - Single verse: `John 1:1` - Intra-chapter range: `John 1:1-5` - Cross-chapter range: `John 1:1-2:11` - Multiple passages (comma- or `+`-separated): `John 1:1-5, Romans 8:1-4` Translate each reference into the shared `{ book, range }` shape used by `query_morphology` and `bible_lookup`. `range` format: `"chapter:verse-chapter:verse"` (e.g., `"1:1-1:5"`). **One book per call.** A cross-book span is split into per-book calls, each with its own `query_morphology` invocation. **Book not found:** If `query_morphology` returns `BOOK_NOT_FOUND`, surface the error verbatim plus any `suggestions` it provides...