ATTY57/toolfault-bench
ToolFault-Bench Can an AI agent tell when a tool result has gone wrong, and does it pick a safe way to recover? ToolFault-Bench contains 522 agent trajectories. Each one ends with a tool result that may be faulty, and the model must answer two questions: fault_type: what, if anything, is wrong with the last tool result (13 classes, including none) next_action: the agent's best next move (11 options) Most tool-use benchmarks measure the happy path. This one covers what happens… See the full description on the dataset page: https://huggingface.co/datasets/ATTY57/toolfault-bench.
| Author / organization | ATTY57 |
|---|---|
| License (as declared) | cc-by-4.0 |
| Downloads | 150 (as reported by Hugging Face) |
| Likes | 0 |
| Created | 24 September 2026 |
| Last updated at source | 24 September 2026 |
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 16 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.