II.
TestSet overview
Reference · livetest-set:the-agent-company-workplace-tasks
TheAgentCompany workplace task suite overview
Canonical TheAgentCompany artifact for workplace software-agent task completion evaluation.
Attributes
displayName
TheAgentCompany workplace task suite
benchmarkId
composition
TheAgentCompany's extensible workplace benchmark task suite asks
agents to browse, code, run programs, communicate with coworkers,
and complete professional digital-worker tasks in a simulated
software-company environment.
homepageUrl
description
Canonical TheAgentCompany artifact for workplace software-agent
task completion evaluation.
Outgoing edges
belongs_to_benchmark1
- benchmark:the-agent-company·BenchmarkTheAgentCompany
Incoming edges
None.