TRAIL

Trace reasoning and agentic issue localization; 148 long-context traces, 841 errors, 20+ error types; Hugging Face dataset.

Stars

24

Adoption surface

mostly simple

Autonomy

n/a

Recovery

n/a

License

✅ open-source

Category

Evaluation and benchmarking harnesses

Repository ↗ Example: TRAIL dataset card ↗

Related in Evaluation and benchmarking harnesses

• Agent Lightning
• SWE-bench
• AgentBench
• inspect_ai
• WebArena
• ClawBench

Maintain TRAIL? Add the badge

Paste this into your README so readers can find the list and see how TRAIL compares:

[![Best of Agent Harnesses](https://img.shields.io/badge/%F0%9F%8F%86_Best_of-Agent_Harnesses-5ac4bf)](https://github.com/RyanAlberts/best-of-Agent-Harnesses)

Copy