grill-with-docs

Stress-test AI-generated plans against domain models and documented decisions.

1|1|Updated Apr 20, 2026
One-click install
npx skills add https://github.com/svssdeva/agentic-skills --skill grill-with-docs-svssdeva
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: grill-with-docs
Source: https://github.com/svssdeva/agentic-skills/tree/main/engineering/grill-with-docs
Command: npx skills add https://github.com/svssdeva/agentic-skills --skill grill-with-docs-svssdeva

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you validate a plan by challenging it against your existing domain model and documented decisions, preventing vague or inconsistent outcomes from slipping through.

Core Features & Use Cases

  • Relentless plan interrogation: Breaks down the design tree by asking one question at a time until shared understanding is reached.
  • Code-and-doc cross-checking: Inspects the codebase when possible and immediately reconciles contradictions with the plan.
  • Domain glossary sharpening: Calls out terminology conflicts and proposes precise canonical terms when language is fuzzy or overloaded.
  • Live documentation updates: Updates CONTEXT.md inline as terms are resolved to keep decisions current and meaningful.

Quick Start

Use the grill-with-docs skill to stress-test my proposed plan against our current domain model and documentation in CONTEXT.md and ADRs.

Frequently Asked Questions about grill-with-docs

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I validate an architecture plan against existing domain documentation?

Validate an architecture plan by stress-testing its assumptions against documented decisions and domain models, interrogating terminology conflicts one question at a time to ensure shared understanding and alignment with established language.

Can I reconcile fuzzy domain terminology with my CONTEXT.md file during design?

Reconcile fuzzy domain terminology with CONTEXT.md by calling out overloaded terms during plan interrogation, proposing precise canonical definitions, and writing inline updates directly into the documentation as terms are resolved.

What is the best way to cross-check an AI-generated plan against a codebase?

Cross-check an AI-generated plan by exploring the codebase to inspect existing dependencies and domain logic, immediately reconciling any contradictions found between the proposed plan and the actual implementation.

How do I ensure architecture decisions align with existing ADRs before implementation?

Ensure decisions align with existing ADRs by interrogating the proposed plan against documented architecture decisions, breaking down the design tree through gated questioning until terminology and logic match established records.

Does plan validation require codebase exploration when answers exist in code?

Plan validation mandates codebase exploration when relevant answers exist in code, ensuring the proposed software design is checked against actual dependencies and implementation logic rather than unverified assumptions.

When should I not use domain model interrogation for software design reviews?

Avoid domain model interrogation when a project lacks established CONTEXT.md or ADRs, as this approach relies on stress-testing plans against existing documented decisions and reconciling terminology with a mature domain glossary.