Paper Detail

ClawGym: A Scalable Framework for Building Effective Claw Agents

Fei Bai, Huatong Song, Shuang Sun, Daixuan Cheng, Yike Yang, Chuan Hao, Renyuan Li, Feng Chang, Yuan Wei, Ran Tao, Bryan Dai, Jian Yang, Wayne Xin Zhao

arxiv Score 16.2

Published 2026-04-29 · First seen 2026-04-30

General AI

Abstract

Claw-style environments support multi-step workflows over local files, tools, and persistent workspace states. However, scalable development around these environments remains constrained by the absence of a systematic framework, especially one for synthesizing verifiable training data and integrating it with agent training and diagnostic evaluation. To address this challenge, we present ClawGym, a scalable framework that supports the full lifecycle of Claw-style personal agent development. Concretely, we construct ClawGym-SynData, a diverse dataset of 13.5K filtered tasks synthesized from persona-driven intents and skill-grounded operations, paired with realistic mock workspaces and hybrid verification mechanisms. We then train a family of capable Claw-style models, termed ClawGym-Agents, through supervised fine-tuning on black-box rollout trajectories, and further explore reinforcement learning via a lightweight pipeline that parallelizes rollouts across per-task sandboxes.To support reliable evaluation, we further construct ClawGym-Bench, a benchmark of 200 instances calibrated through automated filtering and human-LLM review. Relevant resources will be soon released at https://github.com/ClawGym.

Workflow Status

Review status
pending
Role
unreviewed
Read priority
now
Vote
Not set.
Saved
no
Collections
Not filed yet.
Next action
Not filled yet.

Reading Brief

No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.

Why It Surfaced

No ranking explanation is available yet.

Tags

No tags.

BibTeX

@article{bai2026clawgym,
  title = {ClawGym: A Scalable Framework for Building Effective Claw Agents},
  author = {Fei Bai and Huatong Song and Shuang Sun and Daixuan Cheng and Yike Yang and Chuan Hao and Renyuan Li and Feng Chang and Yuan Wei and Ran Tao and Bryan Dai and Jian Yang and Wayne Xin Zhao},
  year = {2026},
  abstract = {Claw-style environments support multi-step workflows over local files, tools, and persistent workspace states. However, scalable development around these environments remains constrained by the absence of a systematic framework, especially one for synthesizing verifiable training data and integrating it with agent training and diagnostic evaluation. To address this challenge, we present ClawGym, a scalable framework that supports the full lifecycle of Claw-style personal agent development. Concr},
  url = {https://arxiv.org/abs/2604.26904},
  keywords = {cs.CL, cs.AI, cs.LG, Claw-style environments, multi-step workflows, scalable development, verifiable training data, agent training, diagnostic evaluation, ClawGym-SynData, persona-driven intents, skill-grounded operations, mock workspaces, hybrid verification mechanisms, ClawGym-Agents, supervised fine-tuning, black-box rollout trajectories, reinforcement learning, lightweight pipeline, per-task sandboxes, ClawGym-Bench, automated filtering, human-LLM review, huggingface daily},
  eprint = {2604.26904},
  archiveprefix = {arXiv},
}

Metadata

{}