Brainstorming — Reframe Before You Build
SkillMediaYou MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Reframes the problem, challenges assumptions, and outputs a design doc that downstream skills can read.
Instructions available. Your AI can read the instructions. Execution depends on the setup they require.
Account requirements not reviewed. Check the skill instructions before use; ahel provides instructions and does not run this skill.
Add ahel to your AI once: Claude, ChatGPT, Cursor, Claude Code or Codex. Then ask it to use this.
Then ask your AI: use the Brainstorming skill
What this skill tells your AI
The instructions your AI receives, as published by myths-labs/muse in skills/toolkit/brainstorming/SKILL.md and read by ahel’s review.
Inspired by gstack's /office-hours. Upgraded to output standardized design docs.
Overview
Don't just ask "what do you want to build?" Ask "what pain are you solving?"
This skill reframes the problem before a line of code is written, then outputs
a design doc that feeds into /sprint downstream skills.
The Process
Phase 1: Reframe the Problem
Before exploring solutions, challenge the framing:
- Listen for pain, not features. User says "add a dark mode toggle." Ask: "What's the actual problem? Is it eye strain? Brand consistency? User requests?"
- Push back on scope. User describes 10 features. Ask: "If you could only ship one this week, which one?"
- Challenge 3 premises. Identify the user's hidden assumptions and question them explicitly.
- "You're assuming users want X. What if they actually want Y?"
- "You're assuming this needs to be real-time. Does it?"
- "You're assuming mobile-first. What if desktop is the real use case?"
Rules:
- One question per message
- Multiple choice when possible
- Don't rush past reframing — this is the highest-leverage phase
Phase 2: Explore Approaches
Once the real problem is clear:
- Generate 3 approaches with different tradeoffs:
- Narrowest wedge — ship in hours, validate the thesis
- Balanced — ship in days, covers core use cases
- Full vision — ship in weeks, comprehensive solution
- Recommend one with clear reasoning
- Estimate effort for each (hours, not story points)
Lead with your recommendation. Explain why.
Phase 3: Present the Design
Once approach is chosen:
- Break into sections of 200-300 words
- After each section, ask: "Does this look right so far?"
- Cover: architecture, components, data flow, error handling, testing
- Apply ETHOS.md: Boil the Lake (do the complete thing), YAGNI (don't add what's not needed)
Phase 4: Output Design Doc
Write a standardized design doc to docs/plans/YYYY-MM-DD-<topic>-design.md:
# [Feature Name] Design Doc
## Problem
[1-2 sentences: the actual pain, not the feature request]
## Reframe
[How the problem was reframed during brainstorming]
## Approach
[Chosen approach and why. Brief mention of alternatives considered.]
## Scope
[What's IN scope and what's explicitly OUT]
## Technical Design
[Architecture, components, data flow, key decisions]
## Acceptance Criteria
[Numbered list of verifiable outcomes]
## Test Strategy
[What to test and how]
## Risks
[Top 2-3 risks and mitigations]
This doc is read by downstream skills:
writing-plansreads it to create implementation planarchitect-agentreads it for architecture reviewcode-reviewer-agentreads Acceptance Criteria to verify completeness
Key Principles
- Reframe first — The best feature is often not what was originally requested
- One question at a time — Don't overwhelm with multiple questions
- Ship the Narrowest Wedge (from ETHOS.md) — Scope tight, validate fast
- Three premises — Always challenge at least 3 hidden assumptions
- Design doc is the handoff — Other skills read it, so make it precise
- Be flexible — Go back and clarify when something doesn't make sense
Signals
- GitHub stars
- 35
- Forks
- 4
- Last commit
- Sep 2026
- Hacker News mentions
- 14
Advanced
- Item type
- skill
- Key
brainstorming-myths-labs- Source
- github.com/myths-labs/muse