3 Agents Run Founder Office Hours
Mentor, Critic, and Operator agents each give advice on your founder problem, then one memo. Why it matters: you need cheer and truth.
- Output
- Text
- Input
- Text
Prompt directory
A prompt library for work, code, research, photos, and more. Every prompt shows what you need to provide, what it produces, and which AI models it was tested with.
Need reusable instructions with versions and install commands? Browse Claude Code skills, Cursor skills, or the full agent skills registry—and see how an AI package manager differs from copy-paste.
Browse by model: Gemini prompts, Claude prompts, and Nano Banana photo prompts. Or start with a use case: LinkedIn headshot prompts or product photography prompts.
Browse by series
Explore
7 prompts
Mentor, Critic, and Operator agents each give advice on your founder problem, then one memo. Why it matters: you need cheer and truth.
Five digger agents research in parallel; one editor merges notes, kills duplicates, and marks fights. Why it matters: raw notes are not a decision.
Four agents take turns: write the spec, build it, try to break it, then fix it. Why it matters: fewer half-done features.
Bull, Bear, Risk, and Judge agents each research one stock with sources, then write one brief. Why it matters: one hot take is easy; four roles catch bias. Research only — not a buy button.
Three agents (Pro, Con, Judge) debate your plan out loud, then lock a final version. Why it matters: stops endless “what if” loops.
Four AI helpers take roles (Dreamer, Doubter, Money, Builder) and argue your startup idea, then give you one clear plan. Why it matters: one brain misses things; four catch more.
Set up a structured multi-agent debate. Why it matters: one brain misses things, and roles catch more.