grill-with-docs
A relentless interview to sharpen a plan or design that records the language and decisions it settles in the repository as it goes. User-invoked.
- License 许可
- MIT
- 最近更新
- 2026-11-13
- Stars
- ★ 1
- Forks
- 0
01 前言 Front Matter
- name
- grill-with-docs
- description
- A relentless interview to sharpen a plan or design that records the language and decisions it settles in the repository as it goes. User-invoked.
- disable-model-invocation
- true
02 资源文件 3
- 参考文档agents/openai.yaml145 B
- 许可证LICENSE1.0 KB
- 参考文档references/source-notes.md1.1 KB
03 SKILL.md 正文
Load the grilling skill and run it. As the interview settles things, record
them where the repository keeps them, in the same session:
- Terms. When an answer names a domain concept, coins a word, or resolves
two words for one thing, load
ubiquitous-languageand follow its protocol (detect, propose, decide, record) against the repository's declared glossary. - Hard-to-reverse decisions. When an answer settles something that would
be expensive to undo, record it through the repository's decision mechanism:
the
adragent in Claude Code, or an ADR in the repository's decision folder elsewhere. Routine choices need no record.
Write each record once the user has confirmed the answer, never on your own recommendation alone. The session ends when the grilling frontier is empty and every settled term and decision is recorded.
04 同仓库更多 Skills
wtf
Re-explain the immediately previous LLM response when it did not land, using plain, precise UK English.
writing-for-agents
Writing documents for agents. Use when creating or editing skills, or modifying AGENTS.md or CLAUDE.md.
technical-writing
Writing developer-facing prose that can be skimmed first and trusted enough to finish — READMEs, guides, tutorials, reference docs, proposals, PR descriptions, release notes. Load before drafting, rewriting, restructuring, or editing technical documents, when a doc reads as a wall of text, claims need receipts, or docs must serve AI agents. Covers reader-first structure, falsifiable claims, docs-as-behavior verification, and agent-readable reference shape. For sentence-level voice or generic AI-shaped prose use clarity; for plain English, lay readers, translation, or ASD-STE100 use simple-english; for recording a learning use expectations; for diagrams use diagrams; for API semantics use api-design.
teach
Teach the user a new skill or concept, within this workspace.
find-skills
Discover and, with authorization, install agent skills from the open skills ecosystem. Use when the user explicitly asks to find or install a skill from the external skills ecosystem, asks whether an installable agent skill exists for a task, or wants to extend agent capabilities through skills.sh or a skill repository. Not for choosing among already-installed local skills or selecting software libraries, developer tools, applications, services, frameworks, or platform primitives; use evaluate-existing-solutions for technology choices.
reduce-system-complexity
Reduce total mechanism in a selected existing system or behavior path while conserving its agreed observable behavior and non-functional guarantees. Use when the user explicitly wants fewer branches, states, dependencies, layers, flags, retries, jobs, adapters, or operational moving parts and needs evidence that complexity was removed rather than relocated. Supports read-only diagnosis and authorized reduction slices. Not for routine cleanup, architecture-investment discovery, module-contract design, physical tree design, new behavior, functionality cuts, speculative rewrites, or Redux/functional reducer functions.