frame
Turn a request into a shared understanding before any plan or code: the intent, what is out of scope, the terms and the questions still open, at the depth the change's mode gives the frame step. Use when the mode reaches frame, when the owner says "grill me" or "grill with docs", and when a term is doing two jobs.
- License 许可
- MIT
- 最近更新
- 2026-10-21
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- frame
- description
- Turn a request into a shared understanding before any plan or code: the intent, what is out of scope, the terms and the questions still open, at the depth the change's mode gives the frame step. Use when the mode reaches frame, when the owner says "grill me" or "grill with docs", and when a term is doing two jobs.
02 资源文件 6
- 参考文档references/greenline.md10.0 KB
- 参考文档references/interview.md5.3 KB
- 参考文档references/prototype.md5.1 KB
- 参考文档references/prototype/LOGIC.md5.9 KB
- 参考文档references/prototype/UI.md6.8 KB
- 参考文档references/research.md1.7 KB
03 SKILL.md 正文
Frame
Frame turns a request into a shared understanding of what is wanted before any plan or code exists: the intent, what is out of scope, the terms, and the questions still open. It runs where the change's mode reaches the frame step, at the depth the mode gives it, and alone when the owner asks for it by name: "grill me", "grill with docs", or a term doing two jobs, which it names in one line with the sharper word it proposes. It also runs mid-work, at any step, the moment scope, intent or a term turns out unclear: that ambiguity is put to the owner at the inline depth, never guessed through. Asked for by name, or loaded by a start, with no change open, it runs at the written depth, its result stays in the reply, and it writes no file. It reads the settled decisions first, in the repository's decision home and the change's record, and reopens none of them. It never plans and never builds: what is built, and in which order, is the plan step's. Facts are yours to look up first, decisions the owner's.
The method
Name the intent, and find the one consequential question the request leaves open, if there is one. A term the request leaves unclear is such a question: put it to the owner, never settle it by your own reading. Ask it with your recommended answer and wait.
Read the repository's documents before the first question: its decision home, its glossary (the repository's own, or GLOSSARY.md where it keeps none) and the records of earlier changes in the same area, so that no question asks what they already settle.
Where the frame interviews the owner, run the interview in references/interview.md to its end: every behaviour the request leaves open goes to the owner, each with your recommended answer, and none is settled on your own.
When asking cannot settle a question
Two kinds of question are answered by work rather than by asking, and each has its page, read only when one arises:
- A question that waits on a fact the repository does not hold, which primary sources must answer: references/research.md.
- A question only trying can answer, whether a logic or state model feels right or what a screen should look like, or a design with no precedent whose shape is not chosen yet: references/prototype.md.
Handoff
Consumes: the request and the change record's front matter (its mode and why); at the written depth, the repository's decision home, its glossary and the records of earlier changes in the same area Produces: the change record's frame section in .greenline/ledger/<id>.md, one line at the inline depth, and at the written depth the confirmed restatement, its out-of-scope list, the settled terms and any open question; the reply alone when no change is open Next: the next step of the change's mode
04 同仓库更多 Skills
use-garden
Consulting garden, the separately installed guidance command this repository's owner enabled, when an open engineering choice the repository does not settle, made in a change or only recommended, would be informed by published guidance. Use for a choice between approaches, a pattern or pitfall for a technology or concern the work touches, or a unit a repository decision names; not for a question about this repository's own facts or a change whose choices are already settled.
use-garden
Consulting garden, the separately installed guidance command this repository's owner enabled, when an open engineering choice the repository does not settle, made in a change or only recommended, would be informed by published guidance. Use for a choice between approaches, a pattern or pitfall for a technology or concern the work touches, or a unit a repository decision names; not for a question about this repository's own facts or a change whose choices are already settled.
what-to-build
Propose what this product could do next at three altitudes: sharpen a shipped slice, extend one, or open ground the product does not stand on yet. Use for 'what should we build next', 'any ideas for this app', 'what's missing', 'is this done or is there more in it', and quick-win hunts. Not for refactors, which belong to the architecture survey.
wayfinder
Plan a huge chunk of work (more than one agent session can hold) as a shared map of decision tickets in a plan directory's map.md, and resolve them one at a time until the way to the destination is clear.
verify
Prove a change's claims with fresh evidence: restate each item of its acceptance so it can fail, capture a baseline and a treatment with the same command, compare the raw artifacts, and return VERIFIED, NOT VERIFIED or INCONCLUSIVE for each, at the depth the change's mode gives the verify step. Use when the mode reaches verify, or when the owner says "verify this", "prove it works" or "did this fix it".
tdd
Test-first development at any step: discover the repository's own test commands, name the break a test catches, watch it fail for the intended reason, make it pass with the smallest change, and add a new test only where it catches a realistic failure the existing ones would miss. Use when writing or changing a test, building a feature or fixing a bug test-first, or deciding whether a change needs a new test; also on 'red-green-refactor' and integration tests.