grill-with-docs-skill

What I do

I orchestrate a focused session that combines two engines:

1. grilling-skill — relentlessly interviews the user about the plan, one question at a time with a recommended answer, until the decision tree is resolved
2. domain-modeling-skill — captures the crystallised terminology into CONTEXT.md and records hard-to-reverse decisions as ADRs, inline as they emerge

The result is alignment AND a durable artifact: the next session (or the next agent) inherits a sharpened glossary and recorded decisions, so the work doesn't have to be re-litigated.

When to use me

Use this skill when:

• You want to stress-test a plan or design before building
• You want the alignment session to ALSO produce domain docs (glossary + ADRs)
• You're starting a non-trivial feature and want to lock down terminology first

Trigger phrases:

• "grill with docs"
• "grill me and write it up"
• "stress-test this plan and capture it"
• "interview me and document"

If you want the interview WITHOUT doc capture, use grill-me-skill instead.

Core Workflow

Step 1: Load both engines

This skill delegates all behavior to two model-invoked engines. Do not reimplement them:

• grilling-skill — owns the interview loop
• domain-modeling-skill — owns doc capture

Step 2: Run the grilling session

Invoke grilling-skill to conduct the interview. Apply its rules throughout:

• One question at a time, each with a recommended answer
• Explore the codebase instead of asking when possible
• Walk the decision tree branch-by-branch

Step 3: Capture docs inline during the session

As grilling-skill resolves each branch, hand off to domain-modeling-skill to:

• Challenge and sharpen terminology against the existing CONTEXT.md
• Update CONTEXT.md inline the moment a term is resolved
• Offer an ADR only when a decision is hard-to-reverse + surprising + a real trade-off

The two engines run interleaved, not sequentially. Doc capture happens during the interview, not after.