Conducts an iterative, challenging interview to refine a technical plan or design while simultaneously generating Architecture Decision Records (ADRs) and a domain glossary.
Quality gate: PASS · better answers · measured 2026-10-23 Blind-judged against the bare model on 2 requests: average lift +7.5 points (specificity, structure, method, fit). won 2/2, lift +7.5, tokens ×89.17
Security scan: clean · 2026-10-24 · 2 files · ran in sandbox no network calls, shell tricks, credential access or instruction tampering found in its files.
Try asking
Say it in your own words; the assistant finds this recipe by itself. Examples:
“Jeg har en udkast til vores nye bestillingsflow. Udfordr mig på valget af database og hjælp mig med at skrive ADR'er for beslutningerne.”
“Gennemgå mit design-dokument for kundeportalen. Stil skarpe spørgsmål om fejlhåndtering og opdater ordbogen med de nye begreber vi definerer.”
“Jeg er usikker på om vi skal bruge event sourcing eller traditionel CRUD. Interview mig for at tvinge en beslutning og dokumentér det i en ADR.”
“I have a draft for our new payment service architecture. Please grill me on the data consistency choices and help me write the ADRs for the decisions we make.”
“Review my design doc for the user onboarding flow. Ask hard questions about edge cases and update the glossary with any new terms we define.”
“I'm stuck on whether to use microservices or a monolith for this feature. Interview me to force a decision and document the rationale in an ADR.”
Try this skill from Claude, ChatGPT or a chat on your own page. Free plan, no credit card.Get started free
Do not use this for simple code reviews, bug fixes, or straightforward implementation tasks that do not involve architectural decisions or domain modeling. It is not suitable for generating initial brainstorming ideas without a concrete plan to critique, nor for pure coding assistance without a design context.
Details
When it is the right choice
Use this skill when you have a draft plan, system design, or feature specification that needs rigorous stress-testing and clarification. It is ideal for sessions where you want to resolve ambiguities, make explicit trade-offs, and produce formal documentation artifacts like ADRs and glossaries as part of the discussion.
The full recipe the assistant reads
Call the Skill tool twice, for "grilling" and "domain-modeling".
Files in this skill (3)
LICENSE
SKILL.md
agents/openai.yaml
Where it comes from
Vendored from mattpocock/skills, commit 5beb35b040e3, path skills/engineering/grill-with-docs, licence MIT. The upstream licence file ships inside the skill.