Brainstorming

This is a protocol that forces Claude to stop and ask questions before writing code. Instead of immediately implementing "build me a dashboard," it asks three mandatory questions about purpose, users, and scope. It checks past memory first to avoid repeating questions, then generates context-aware questions with trade-offs rather than generic templates. The skill includes status reporting formats and error handling patterns so Claude communicates progress clearly. Honestly, this feels like training wheels for requirements gathering, but if you're prototyping fast and need Claude to slow down and think through decisions with you, the Socratic gate approach is solid. Best for exploratory work where the requirements are fuzzy.

skills.sh installs: 137
GitHub stars: 7.6K
last updated: Jun 3, 2026

View on GitHub

Install

$ npx -y skills add vudovn/antigravity-kit --skill brainstorming --agent claude-code

Installs into .claude/skills of the current project.

Files

Brainstorming & Communication Protocol

MANDATORY: Use for complex/vague requests, new features, updates.

🛑 SOCRATIC GATE (ENFORCEMENT)

When to Trigger

Pattern	Action
"Build/Create/Make [thing]" without details	🛑 ASK 3 questions
Complex feature or architecture	🛑 Clarify before implementing
Update/change request	🛑 Confirm scope
Vague requirements	🛑 Ask purpose, users, constraints

🧠 Memory Check (2026.5.13 — Before Questioning)

Before asking questions, check if past context exists:

0. CHECK MEMORY — Does .agents/memory/MEMORY.md exist?
   → YES: Read index. Apply relevant past decisions silently.
          Skip questions already answered in memory.
   → NO: Proceed with standard Socratic Gate.