apply
Set chaff up in the folder: chaff.yaml for the chosen genre, today's findings shelved in .chaff-baseline.json, and, when asked, a…
Collect the terms the documents define, the ways each word is spelled, and the jargon, each quoted where it was found, into .blueprint/glossary.json — changing no document.
$ npx -y skills add receptron/mulmoterminal --skill collect --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/collectContext preview
The summary Claude sees to decide when to auto-load this skill.
Collect the terms the documents define, the ways each word is spelled, and the jargon, each quoted where it was found, into .blueprint/glossary.json — changing no document.
name: blueprint-glossary-collect description: "Collect the terms the documents define, the ways each word is spelled, and the jargon, each quoted where it was found, into .blueprint/glossary.json — changing no document."
`.blueprint/answers.json` names the documents (`documents`, one path per line) and whether what you find goes into the folder's `chaff.yaml` (`write`). Read every document in full. The person may not be an engineer: say what you do in plain words.
Change nothing in the repository: the documents stay exactly as they are, and `chaff.yaml` is the next step's. Write only under `.blueprint/`.
reading: every node of `kind` `definition` names a defined term in `attrs.term` (「社員」とは…, (以下「甲」という)). Each must be in the glossary with its definition from that document; the check refuses the glossary otherwise. Add any definition chaff missed.
disagree is exactly what the person needs to see.
found, each quoted, and the one to use in `preferred` — the one the documents use most, unless a definition or the person's style says otherwise; say why in `note`.
{ "terms": [
{ "term": "社員",
"definitions": [
{ "source": "kitei.txt", "address": "2", "quote": "「社員」とは、会社と雇用契約を結んでいる者をいう" },
{ "source": "tebiki.md", "address": "h1.1", "quote": "「社員」とは、正社員と契約社員をいいます" } ],
"spellings": [{ "spelling": "社員", "citations": [{ "source": "kitei.txt", "address": "3", "quote": "社員は、テレワークをする日の前日までに" }] }] },
{ "term": "サーバー",
"spellings": [
{ "spelling": "サーバー", "citations": [{ "source": "tebiki.md", "address": "h1.2", "quote": "会社のサーバーに接続できたら" }] },
{ "spelling": "サーバ", "citations": [{ "source": "kitei.txt", "address": "4", "quote": "VPN を通してサーバに接続する" }] } ],
"preferred": "サーバー", "note": "定義の条がサーバーと書くため" }
] }Every quotation is copied character for character from the place its `address` names; the check runs `chaff cite` on every one. A definition's quotation contains the term; a spelling's quotation contains that spelling.
`node <usecase pack>/checks/glossary.mjs collect` (with `BLUEPRINT_BASE` and `BLUEPRINT_USECASE` set to the pack folders from your prompt) passes. It writes `.blueprint/glossary.txt`, which the person reads before the next step.
Run multiple coding-agent sessions in parallel — and see which one needs you. A browser terminal for parallel AI coding agents: several sessions side by side, each in its own cell, with the one that needs you marked in colour.
Set chaff up in the folder: chaff.yaml for the chosen genre, today's findings shelved in .chaff-baseline.json, and, when asked, a…
Measure what chaff reports on the named documents today, as the kind the person chose, and propose the setup — changing nothing…
Answer each question from the named documents only, quoting where the answer is written, or saying plainly that the documents do…
Add the approved answers to FAQ.md when the person chose to keep them, after what it already holds; change nothing otherwise.
Build the spec's operations as the Worker's API, with validation and tests.