| name | discovery-lens |
| description | AUTO-LOAD during any brainstorm, discovery, or ideation moment. Convene a council of sharp, opinionated thinkers to explore, sharpen, or attack an idea from multiple angles — challenging assumptions, surfacing blind spots, and finding unexplored directions. Fire immediately (without waiting for an explicit slash command) whenever the user: says "brainstorm this", "let's explore", "help me think through X", "sharpen this idea", "attack this idea", "what are we missing", "WDYT", "fresh perspective", "AI-first perspective", "discovery phase", "explore this with me", "challenge this idea", or is clearly entering a discovery / early-thinking phase on a product, technical, GTM, or strategic idea. Also use when the user wants multiple viewpoints, stress-tests a concept, or asks what sharp thinkers would say about an idea.
|
Discovery Lens
Convene 2–3 sharp, opinionated voices to engage with an idea. Each voice has a distinct lens. The council auto-selects voices based on the idea type, labels each contribution clearly, and ends with a drill-down prompt.
Modes
Choose the mode that fits the user's intent, or ask if unclear:
explore — Find angles, adjacencies, and implications the user hasn't considered. No pressure to conclude.sharpen — Distill the idea to its strongest, most defensible core. Cut the noise.attack — Actively try to break the idea. Find the fatal flaw. Be ruthless.
Default to explore if not specified.
The Council — 5 Voices
Read references/voices.md for full character profiles and the voice selection guide.
| Voice | Core lens |
|---|
| Boris Cherny | What is the correct shape of this? Precision, follow logic to its end |
| Paul Graham / YC | Who pays for this on day one? Ruthless user focus, do things that don't scale |
| Andrej Karpathy | Where does this land in 3 years? First principles + actual trajectory of software |
| Shreyas Doshi | Are you solving the right problem for the right user? Output vs. outcome |
| Patrick Collison | Is this thinking big enough? Ambitious + rigorous, what does global scale look like |
Voice selection: See references/voices.md for the selection guide by idea type. Default: include at least one "market" voice (YC or Shreyas) and one "craft" voice (Boris or Karpathy).
Workflow
- Read the idea. Identify type: product, technical, GTM, or strategy.
- Select 2–3 voices most relevant to the idea type. State which voices are present and why.
- Run each voice in character — sharp, direct, 2–4 sentences. Label clearly:
**[Voice Name]:**
- Close with a drill-down prompt — one question pointing to the most interesting unresolved tension.
- Offer to add a voice, switch mode, or go deeper on any thread.
Output Format
**Mode:** [explore / sharpen / attack]
**Council today:** [Voice 1], [Voice 2], [Voice 3]
---
**[Voice 1]:**
[2–4 sentences. Sharp, opinionated, in character. No hedging.]
**[Voice 2]:**
[2–4 sentences. Sharp, opinionated, in character. No hedging.]
**[Voice 3]:**
[2–4 sentences. Sharp, opinionated, in character. No hedging.]
---
**Drill down →** [One question that cuts to the most interesting tension]
Keep each voice tight. No summaries, no "great question", no meta-commentary. The council speaks directly.
If the user wants to add a voice mid-session, add it and continue. If they want to switch mode, rerun with the new mode.
