{
  "id": "test-set:agentbench-environments",
  "_kind": "TestSet",
  "_file": "benchmarks/test-sets/source-backed-benchmark-test-sets.yaml",
  "_cluster": "benchmarks",
  "attributes": {
    "displayName": "AgentBench multi-environment suite",
    "benchmarkId": "benchmark:agentbench",
    "environmentCount": 8,
    "releasedAt": "2023-08-07",
    "composition": "Multi-environment LLM-as-agent benchmark covering eight interactive\nenvironments for reasoning, decision-making, tool use, and instruction\nfollowing.\n",
    "homepageUrl": "https://github.com/THUDM/AgentBench",
    "description": "Canonical AgentBench artifact for broad LLM-as-agent evaluation.\n"
  },
  "outgoingEdges": [
    {
      "from": "test-set:agentbench-environments",
      "to": "benchmark:agentbench",
      "kind": "belongs_to_benchmark"
    }
  ],
  "incomingEdges": []
}