Act on feedback with judgment, fixing only verified bugs

Reads feedback from wherever it lives, sorts each item by what it asks, checks whether errors were really seen, fixes the verified bugs and proposes design changes with restraint. For anyone who receives review notes or bug reports on their work.

What it asks for

  • about (Optional, words): What this run is for, in the person's words.
  • known (Optional, words): What is already known, decided or done, so the run starts from it instead of asking again.

The steps your AI walks

  1. Read the feedback where it lives
  2. Sort each item by what it asks
  3. Check whether the error was seen for real
  4. Verify and fix the clear bugs
  5. Propose changes to how it looks or works, with restraint
  6. Verify what changed and report short

Use it from your AI

Connect the AI you already use to Trillion once. Then ask in your own words; your AI finds this workflow by what it does and runs it.

Say to your AI: “Use Trillion to reads feedback from wherever it lives, sorts each item by what it asks, checks whether errors were really seen, fixes the verified bugs and proposes design changes with restraint. For anyone who receives review notes or bug reports on their work”