papers / hundred-step

Hundred Step: Dependency Preservation and Recovery in Long Workflows

Pooria Arab · orcid 0009-0005-9964-6841 · 2026-10-29 · Published on Zenodo

abstract

Long workflows require agents to retain exact dependencies and recover from faults without skipping work. Hundred Step isolates these behaviors in 24 synthetic workflows with replay-based scoring. A reactive scripted controller completes all tasks: 100.0% (95% Wilson interval 86.2 to 100.0). Disabling recovery or allowing transient retries alone prevents completion on every task. We also execute 4608 simulated episodes with matched omission draws and a recovery ablation. The stipulated omission probability is 0.02 per logical step. With recovery enabled, simulated completion is 60.4% (95% seed-bootstrap interval 53.5 to 67.4) at 25 steps. Simulated completion is 13.2% (8.3 to 18.8) at 100 steps. The study validates diagnostic behavior under explicit controls. It provides no learned-model leaderboard, human baseline, or estimate of production reliability.

benchmark: benchmark-hundred-step. submitted to: none recorded.

citation

Arab, Pooria (2026). Hundred Step: Dependency Preservation and Recovery in Long Workflows. https://doi.org/10.5281/zenodo.23752401

bibtex

@misc{hundred-step,
  author = {Arab, Pooria},
  title = {Hundred Step: Dependency Preservation and Recovery in Long Workflows},
  year = {2026},
  doi = {10.5281/zenodo.23752401},
  url = {https://pooriaarab.com/papers/hundred-step},
}

venues, counts and the software records live on papers.