Agent Evaluation Reporting
Use when summarizing agent evaluations where autonomous, assisted, failed, timed-out, or invalid outcomes must remain distinct and comparable.
4,595,173 AI resources indexedWorking toward the world’s largest AI resource index.
Reusable instructions for compatible AI agents.
Field counts come from the latest full-catalog snapshot. A resource can belong to several fields and platforms.
Use when summarizing agent evaluations where autonomous, assisted, failed, timed-out, or invalid outcomes must remain distinct and comparable.
Use when an agent workflow needs deterministic recovery evidence for sandbox, MCP/tool, worker, checkpoint, memory, or orchestration failures.
Use when auditing whether a field, enum, flag, or API contract propagates consistently across storage, services, clients, analytics, and tests.
Try a broader search or clear some filters.
Platform labels describe intended use. Read each resource's author credits, requirements and available access methods.