Looking for alternatives to get-shit-done in Agent frameworks? Here are 9 comparable open-source projects, with our health scores and licenses so you can choose with confidence.
claw-code is an open-source Rust project in the agent frameworks space with 193.4k GitHub stars. The repository is actively maintained, with recent commits.
hermes-agent is an open-source Python project in the agent frameworks space with 185.8k GitHub stars. The repository is actively maintained, with recent commits.
AutoGPT is a platform for building, deploying, and running continuous AI agents that automate multi-step workflows. It pairs a low-code visual agent builder (connecting "blocks" into workflows) with a server runtime, marketplace of pre-built agents, and the original standalone AutoGPT Agent plus the Forge toolkit and agbenchmark.
LobeHub (formerly LobeChat) is an open-source, self-hostable platform for building and operating teams of AI agents. It centers on "agents as the unit of work," offering an agent builder, multi-agent collaboration (groups, pages, scheduling, projects), MCP-compatible plugins/tools, persistent personal memory, and support for many model providers.
memos is an open-source Go project with 60.7k GitHub stars. It is associated with docker, foss, go, markdown. The repository is actively maintained, with recent commits.
okhttp is an open-source Kotlin project with 47k GitHub stars. It is associated with android, graalvm, java, kotlin. The repository is actively maintained, with recent commits.
curl is an open-source C project with 42.1k GitHub stars. It is associated with c, client, curl, ftp. The repository is actively maintained, with recent commits.
tidb is an open-source Go project with 40.2k GitHub stars. It is associated with agent, agent-context, agent-memory, agentic. The repository is actively maintained, with recent commits.
Minds Platform (Minds-Cowork) is an open-source, general-purpose AI agent application aimed at knowledge workers, packaged as a web/desktop (Electron) app that automates multi-step read-and-write tasks and lets users build internal AI tools without engineering. It emphasizes self-hosted control with deployment across cloud, VPC, on-prem, and air-gapped environments.
See the full get-shit-done overview or browse all Agent frameworks.