II.
Benchmark overview
Reference · livebenchmark:terminal-bench
Terminal-Bench overview
Benchmark for AI agents performing real software-engineering and devops tasks in a terminal environment.
Attributes
displayName
Terminal-Bench
homepageUrl
kind
terminal-agent
targetsKind
AgentVersion
description
Benchmark for AI agents performing real software-engineering and devops tasks in a terminal environment.
Outgoing edges
covers1
- skill-area:cli-design·SkillAreaCLI Design
Incoming edges
belongs_to_benchmark1
- test-set:terminal-bench-v1·TestSetTerminal-Bench v1
bounds_subject1
- scope-boundary:terminal-bench.scope·ScopeBoundary
for_benchmark1
scored_against1