Agent harnesses, runtimes, and frameworks for building and operating Claw-class agents.
8 resources
An agent-team orchestrator for one-person companies, formerly named TinyClaw; it is cataloged under its current name to avoid confusing it with the separate TinyClaw project.
An early Claw-branded project from Xpose Market whose public repository does not yet document enough persistent-agent behavior for inclusion in the verified Claw directory.
An experimental minimal Rust implementation inspired by OpenClaw, listed as an early runtime project while its agent capabilities receive further verification.
An experimental C rewrite of OpenClaw targeting a very small static binary and fast startup; the public repository currently provides limited documentation.
A continual-learning layer that lets compatible Claw agents improve skills and behavior from experience through memory, scheduled training, and asynchronous reinforcement learning.
OpenSpace by HKUDS (University of Hong Kong) is an open-source self-evolving skill engine that empowers AI Agents to automatically learn, reuse, and optimize skills from every task, making automation smarter and significantly more token-efficient.
DeerFlow 2.0 is ByteDance's open-source SuperAgent harness (Deep Exploration and Efficient Research Flow) that orchestrates sub-agents, persistent memory, Docker sandboxes, and on-demand skills to autonomously handle complex, multi-hour tasks like deep research, coding, data analysis, and full content creation.
ClawTeam turns solo AI agents into intelligent swarms that self-organize, divide labor, and deliver full automation from a single human command.
All ClawAll Claw
