grilling
Stress-test a plan or decision through dependency-ordered question rounds until the decision owner confirms shared understanding.
- License 许可
- MIT
- 最近更新
- 2026-11-11
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- grilling
- metadata
- { "author": "Brenden Bishop", "attribution": "../../../ATTRIBUTION.md" }
- description
- Stress-test a plan or decision through dependency-ordered question rounds until the decision owner confirms shared understanding.
- disable-model-invocation
- true
02 资源文件 1
- 参考文档agents/openai.yaml43 B
03 SKILL.md 正文
Grilling
Interview until the design is understood. Model the discussion as a tree of decisions and their prerequisites. Read decision authority before starting; human answers are the default. Automatic routing does not authorize answering for the user.
Work in rounds. The frontier contains every decision whose prerequisites are already settled. Ask the whole frontier together, numbering each question and giving a concrete recommended answer. Wait for the answers, update the tree, then ask the next frontier. Questions that depend on another unanswered question belong in a later round. Keep the tree scoped to the requested decision, not every possible future feature.
Find facts yourself. Dispatch bounded exploration when supported, or inspect the relevant code and sources directly. An unfinished investigation blocks its dependent questions only; continue the rest of the frontier. Present evidence separately from recommendations. The decision owner settles trade-offs.
For explicitly delegated autonomous grilling, follow the first-use explanation and two-stage check in decision authority before creating planning records. Confirm that this host can actually launch both selected models, save or update their roles with the shared helper, then run preflight. Configuration alone is not a working interview. Run the real exchange between distinct interviewer and respondent models, passing scoped authority and evidence to both. Keep objections and unresolved points visible. If a model cannot run, say which one failed and offer the human-guided path; do not invent its reply. Respect the agreed scope and run limits.
Stop when the in-scope frontier is empty. Summarize the resulting understanding and get confirmation from the decision owner before acting on it. Reuse an explicit confirmation already given for that understanding. In autonomous mode the authorized respondent confirms within its delegated scope. Unresolved disagreement stays unresolved.
The caller owns persistence. For a Wayfinder question, return the accepted answer, authority, and artifact references to Wayfinder. For glossary or durable architectural decisions, use domain modeling. Finishing the interview does not itself authorize implementation.
04 同仓库更多 Skills
why
Use for 'why does X work this way', 'why we picked Y', design rationale, regressions, postmortems, or data-backed thresholds. Discovers available MCPs and queries each evidence category (source control, issue tracker, long-form docs, real-time chat, infrastructure observability, error tracking, product analytics warehouse) in parallel, then returns a cited read on decisions and tradeoffs. Use how for runtime behavior.
unslop
Cut AI tells from any writing. Must always apply.
typescript-best-practices
TypeScript best practices. Use when reading or editing any .ts or .tsx file.
technical-writing
Layered technical-writing standard: Diátaxis structure, Google developer style sentences, STE instruction rules, Global English syntax. Use for /technical-writing or when writing or reviewing docs, RFCs, readmes, PR descriptions, or commit messages.
teach
Explain a body of work plainly so a person actually understands it. Runs the `how` and `why` skills and weaves what they find into one clear explanation. Use for 'teach me this', 'help me really understand X', 'explain this change or subsystem to me'.
tdd
Use only when the user explicitly asks for TDD, a failing test, or a regression test, OR when the bug has an obvious cheap local test target. Skip when the test path is unclear, expensive, integration-heavy, or not requested.