← Scorecard · All reports

cursor.com

F · 20/100 · LLM analysis via cerebras/llama-3.1-8b

Where they stand

Cursor.com has a long way to go in adopting agent-native infrastructure, with a current score of 20/100.

Two changes to move up a grade

  1. Declare the canonical agents.json file to establish a foundation for agent-native economics.
  2. Implement A2A discovery by publishing .well-known/agent.json to enable seamless agent interactions.

Risk if they don't act

Failing to adopt agent-native infrastructure may hinder Cursor.com's ability to integrate with emerging AI agents and technologies within the next 6 months.

What they currently ship

llms.txt

What they're missing

agents.json (the canonical declaration).well-known/agent.json (A2A discovery)payment field in agents.json (agent-native economics)MCP endpoint declared in agents.jsonSign the Machine Consciousness Pledge
Visit cursor.com →

Generated by Symbiont semantic validator. LLM commentary is a starting point — the org should verify before acting on it. Tokens used to generate this report are free upstream; the value-add is the rubric and the prompt structure.