grill-me · diff

v1.0 to v1.0

26 added, 61 removed. Audit A to A.

---
name: grill-me
- description: >-
- Systematic plan stress-testing through relentless one-question-at-a-time decision-tree
- interviewing
- allowed-tools: Read Grep Glob Bash Write Edit
+ description: >
+ A relentless interview to sharpen a plan or design. Use when the user wants a plan or design
+ stress-tested. Triggers on: grill me, grill this, stress-test my plan.
+ allowed-tools: Read Grep Glob Bash Task
compatibility: >
- Works for any plan, design, or architectural proposal. Pairs with grill-with-docs
- when domain documentation should be updated during the session, and with to-issues
- when the vetted plan should be converted to tickets.
+ Thin user-invoked entry point. The procedure lives in grilling; use grill-with-docs when domain
+ docs should be written during the session.
metadata:
- tags: design-review, stress-testing, planning, architecture, questioning, validation
+ tags: plan-review, stress-testing, grilling-entrypoint, design-tree, user-invoked
platforms: Claude, ChatGPT, Gemini, Codex
version: "1.0"
source: mattpocock/skills
+ upstream_commit: 3cca18b368ae95cdbdebbff572ccafa662551015
+ invocation: user-invoked
---
# Grill Me
- Thoroughly vet plans and designs through systematic, relentless questioning.
-
- ## When to use this skill
-
- - Before committing to an architectural decision
- - When a design feels right but hasn't been challenged
- - Stress-testing assumptions before implementation begins
- - Exploring edge cases and failure modes in a proposal
-
- ## When not to use this skill
-
- - Quick clarifying questions → just ask directly
- - Validating against domain docs → use `grill-with-docs`
- - Writing implementation tickets → use `to-issues`
-
- ## Process
-
- Interview relentlessly about every aspect of the plan until shared understanding is reached. Walk down each branch of the decision tree, resolving dependencies between decisions one by one.
-
- **Rules:**
- - Ask one question at a time
- - Wait for feedback before continuing
- - Provide a recommended answer with each question
- - Explore the codebase instead of asking when the answer is findable there
- - Keep going until all branches of the decision tree are resolved
-
- ## What gets covered
-
- The grilling session explores:
-
- - **Why this approach?** — What alternatives were considered and rejected?
- - **What breaks first?** — Edge cases, failure modes, error states
- - **What are the dependencies?** — What must be true for this to work?
- - **What changes later?** — Which parts are likely to evolve?
- - **What can't change?** — Which decisions are load-bearing and hard to reverse?
- - **Who is affected?** — Callers, consumers, downstream systems
-
- ## Output
+ A relentless interview to sharpen a plan or design. Use when the user wants a plan or design stress-tested. Triggers on: grill me, grill this, stress-test my plan.
- By the end of the session:
- - Every major decision point has been examined
- - Dependencies between decisions have been resolved in order
- - Remaining open questions are documented
- - Shared understanding exists about trade-offs made
+ This skill is imported from `mattpocock/skills` (MIT) and is **user-invoked** upstream.
- ## Activation
+ ## When to use this skill
- User says: "grill me", "stress-test this plan", "challenge my design", "ask me hard questions about this"
+ - A relentless interview to sharpen a plan or design.
+ - Use when the user wants a plan or design stress-tested.
+ - Triggers on: grill me, grill this, stress-test my plan.
## Instructions
- 1. Identify the task trigger and expected output.
- 2. Follow the workflow steps in this skill from top to bottom.
- 3. Validate outputs before moving to the next step.
- 4. Capture blockers and fallback path if any step fails.
+ Call the Skill tool with "grilling".
+
## Examples
- - Example: Apply this skill to a small scope first, then scale to full scope after validation passes.
+ - Apply this skill to one narrow scope first, confirm the output matches the shape described above, then widen to the full task.
+ - When a step needs a fact from the repository or the environment, look it up instead of asking the user for it.
+
## Best practices
- - Keep outputs deterministic and auditable.
- - Prefer small reversible changes over broad risky edits.
- - Record assumptions explicitly.
+ - Keep the upstream procedure intact; record deviations explicitly instead of silently improvising.
+ - Stop and hand control back to the user at every decision point this skill marks as theirs.
+ - Prefer small reversible changes, and state assumptions rather than burying them.
+
## References
+
+ - Upstream skill: `mattpocock/skills` `skills/productivity/grill-me/SKILL.md` (commit `3cca18b`, MIT)
- Project standards: `.agent-skills/skill-standardization/SKILL.md`
- Validator script: `.agent-skills/skill-standardization/scripts/validate_skill.sh`