The conventional answer to a brief is usually correct and always taken. This finds the unconventional approaches, plans the experiment that would prove each one, and ranks them by effort against payoff - so you can pick a strange route on evidence rather than on nerve.

Execution

  1. brief-analyzer
  2. door-finder
  3. experiment-planner
  4. execution-crafter
  5. approach-ranker

The team

Agent

brief-analyzer

Finds the conventional solution and why it dominates

@cf/openai/gpt-oss-120b

Agent

door-finder

Finds the unconventional approaches

@cf/openai/gpt-oss-120b

Agent

experiment-planner

Plans the cheapest test that would prove each approach

@cf/openai/gpt-oss-120b

Agent

execution-crafter

Details the execution for the surviving approaches

@cf/openai/gpt-oss-120b

Agent

approach-ranker

Ranks the approaches by effort against payoff

@cf/openai/gpt-oss-120b