word-studylisted

# Word study Work under `bible-evidence-grounding`. The empty-result protocol there is not optional decoration in this skill — it is the point of it. ## The sequence, and why step 3 is not skippable **identify the lexeme → lexical entry → usage across scripture → only then say what it means.** A dictionary gloss tells you the range a word *could* cover. Usage tells you what it does cover, for this author, in this period. Every error in `references/exegetical-fallacies.md` — the etymological fallacy, illegitimate totality transfer, the root fallacy, semantic anachronism, cognate overload — is committed by someone who read the lexicon entry and stopped. **Read `references/exegetical-fallacies.md` before writing up any word study.** ### 1. Identify the lexeme If the user gave a **verse**, use `get_interlinear` — it returns each word with its Strong's number and parsing code. If the user gave an **English word**, use `search_original_word` and pick from what comes back. **Never pass the `language` filter to `search_original_word`.** It is broken in production: with the filter set, every query returns zero results. Verified — `love` alone returns 20 entries including G0025 *agapaō*, G0026 *agapē*, G0027 *agapētos*; `love` with `language: "greek"` returns `No lexicon entries found for "love".` Omit the parameter and read the language label on each result. If you pass it, the tool will tell you a word does not exist when it plainly does. ### 2. Clean the number, then get the