v2.3 · Scenario Library and Benchmark Harness

AgentOps Benchmark Console

Reusable deterministic scenarios for governance, policy, runtime, routing, safety, and accelerator control-plane behavior.

8

Packaged scenarios

3

Benchmark suites

0

Live providers/connectors enabled

Core Flow

Scenario → simulated result → decision alignment → runtime alignment → safety alignment → control coverage → evidence coverage → benchmark report.

API

GET /benchmarks/scenarios

GET /benchmarks/suites

POST /benchmarks/run

GET /benchmarks/summary

Boundary

Benchmarks are simulation-only. They validate control-plane behavior, not live LLM quality or live enterprise connector execution.