Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2605.23311 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866913155097034752 |
|---|---|
| author | Yang, Ke Li, Panpan Wu, Zonghan Xu, Kejin Huang, Huaxi Huang, Xiaoshui |
| author_facet | Yang, Ke Li, Panpan Wu, Zonghan Xu, Kejin Huang, Huaxi Huang, Xiaoshui |
| contents | When a structured tool agent fails mid-execution, the runtime faces a dilemma: replaying the entire task is safe but wasteful, while restoring from a local checkpoint is efficient but can leave committed downstream work tied to an upstream history that no longer exists. This tension is acute in commitment-sensitive settings, where rollback targets a single failed instance yet downstream consumers have already acted on its output. Existing recovery approaches provide mechanical rollback but no criterion for whether a local restore remains semantically valid after downstream commitment. We formalize this gap as semantic recoverability and address it in DART, a modular runtime that localizes the failed instance, certifies semantically recoverable boundaries of that instance, aligns checkpoints to those boundaries, and selects an admissible restore point that preserves committed downstream work under dependency and effect constraints-or blocks otherwise. Across three LLM-driven domains and external validation on a LangGraph-based substrate, DART correctly recovers all evaluated commitment-sensitive cases where baseline local recovery fails, and a five-domain safety audit finds no unsafe admitted rollbacks. These results show that controller legality does not imply semantic validity, and that sound local recovery requires an explicit admissibility check. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_23311 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | DART: Semantic Recoverability for Structured Tool Agents Yang, Ke Li, Panpan Wu, Zonghan Xu, Kejin Huang, Huaxi Huang, Xiaoshui Artificial Intelligence When a structured tool agent fails mid-execution, the runtime faces a dilemma: replaying the entire task is safe but wasteful, while restoring from a local checkpoint is efficient but can leave committed downstream work tied to an upstream history that no longer exists. This tension is acute in commitment-sensitive settings, where rollback targets a single failed instance yet downstream consumers have already acted on its output. Existing recovery approaches provide mechanical rollback but no criterion for whether a local restore remains semantically valid after downstream commitment. We formalize this gap as semantic recoverability and address it in DART, a modular runtime that localizes the failed instance, certifies semantically recoverable boundaries of that instance, aligns checkpoints to those boundaries, and selects an admissible restore point that preserves committed downstream work under dependency and effect constraints-or blocks otherwise. Across three LLM-driven domains and external validation on a LangGraph-based substrate, DART correctly recovers all evaluated commitment-sensitive cases where baseline local recovery fails, and a five-domain safety audit finds no unsafe admitted rollbacks. These results show that controller legality does not imply semantic validity, and that sound local recovery requires an explicit admissibility check. |
| title | DART: Semantic Recoverability for Structured Tool Agents |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2605.23311 |