iiRecord
Agentic AI Atlas · OSWorld 369 real-computer tasks
test-set:osworld-369-computer-tasksa5c.ai
II.
TestSet JSON

test-set:osworld-369-computer-tasks

Structured · live

OSWorld 369 real-computer tasks json

Inspect the normalized record payload exactly as the atlas UI reads it.

File · benchmarks/test-sets/source-backed-benchmark-test-sets.yamlCluster · benchmarks
Record JSON
{
  "id": "test-set:osworld-369-computer-tasks",
  "_kind": "TestSet",
  "_file": "benchmarks/test-sets/source-backed-benchmark-test-sets.yaml",
  "_cluster": "benchmarks",
  "attributes": {
    "displayName": "OSWorld 369 real-computer tasks",
    "benchmarkId": "benchmark:os-world",
    "caseCount": 369,
    "composition": "Source-backed OSWorld task suite of real-world desktop and web\ncomputer-use tasks with reproducible setup and execution-based\nevaluation across operating-system and productivity workflows.\n",
    "homepageUrl": "https://os-world.github.io/",
    "description": "Canonical OSWorld artifact for multimodal desktop-computer control\nagent evaluation.\n"
  },
  "outgoingEdges": [
    {
      "from": "test-set:osworld-369-computer-tasks",
      "to": "benchmark:os-world",
      "kind": "belongs_to_benchmark"
    }
  ],
  "incomingEdges": []
}