Demo benchmark — simulated data
Configure an agent readiness benchmark
Select comparable products, define one explicit task, and run the same seeded test conditions across three provider-compatible demo adapters.
1 · Products2 · Workflow3 · Configuration
What product do you want to benchmark?
Search the curated catalog. Suggestions come from validated category relationships, not an LLM guess.
What should the agents accomplish?
Every product receives the same objective, success criteria, viewport, providers, repetition count and action limit.
Run configuration
This MVP runs deterministic demo adapters only. No OpenAI, Claude or Gemini browser API is called.