| name | grill-with-docs |
| description | Grilling session that challenges your plan against the existing domain model, sharpens terminology, and writes agent-readable decisions to design/decisions/ as they crystallise. Use when user wants to stress-test a plan against their project's language and documented decisions. |
| disable-model-invocation | true |
Interview me relentlessly about every aspect of this plan until we reach a shared understanding. Walk down each branch of the design tree, resolving dependencies between decisions one-by-one. For each question, provide your recommended answer.
Ask the questions one at a time, waiting for feedback on each question before continuing.
If a question can be answered by exploring the codebase, explore the codebase instead.
Domain awareness
During codebase exploration, also look for existing documentation:
File structure
Agent-readable decisions live in design/decisions/:
/
├── AGENTS.md <- agent entry point (auto-loaded)
├── ARCHITECTURE.md <- how things connect (lazy-loaded)
├── design/
│ ├── decisions/ <- agent-readable decisions (this skill writes here)
│ │ ├── INDEX.md <- auto-generated scope index
│ │ ├── 0001-slug.md
│ │ └── 0002-slug.md
│ └── lessons.md <- operational learnings (auto-populated via trailers)
└── docs/
└── decisions.md <- human-readable decision narratives (manual, separate)
The design/decisions/ directory contains short, agent-optimised decision records. The human-readable docs/decisions.md is a separate document maintained by the author when a decision is worth the full persuasion narrative. They are related but independent - no sync required.
Two capture paths for decisions
- This skill writes
design/decisions/NNNN-slug.md directly during the grilling session (primary path for architectural decisions).
Codified-Decision: commit trailer on a PR triggers a post-merge workflow that creates the file automatically (same pattern as Codified-Learning: trailers for design/lessons.md).
Both paths produce the same file format in the same directory. Use the trailer when a decision crystallises during a PR rather than a grilling session.
During the session
Sharpen fuzzy language
When the user uses vague or overloaded terms, propose a precise canonical term. "You're saying 'account' - do you mean the Customer or the User? Those are different things."
Discuss concrete scenarios
When domain relationships are being discussed, stress-test them with specific scenarios. Invent scenarios that probe edge cases and force the user to be precise about the boundaries between concepts.
Cross-reference with code
When the user states how something works, check whether the code agrees. If you find a contradiction, surface it: "Your code cancels entire Orders, but you just said partial cancellation is possible - which is right?"
Write decisions to design/decisions/
When a decision crystallises during the session, write it to design/decisions/NNNN-slug.md immediately. Use the format in ADR-FORMAT.md.
After writing or updating decision files, regenerate design/decisions/INDEX.md - a scope-indexed table that agents read to decide which decisions to lazy-load.
Offer decisions sparingly
Only offer to create a decision record when all three are true:
- Hard to reverse - the cost of changing your mind later is meaningful
- Surprising without context - a future reader will wonder "why did they do it this way?"
- The result of a real trade-off - there were genuine alternatives and you picked one for specific reasons
If any of the three is missing, skip the decision. Use the format in ADR-FORMAT.md.
