The open agent registry.

OpenClaw

TYPESCRIPT · v2026.3.7

Any Claws model — defaults to claude-sonnet-4-6

48.2k stars · 1.2M deploys

CodingFlagship

Flagship coding agent. Skills, memory, tool calling and a gateway that speaks OpenAI-compat.

NanoClaw

TYPESCRIPT · v2026.3.4

Any Claws model — defaults to claude-haiku-4-2

6.4k stars · 48k deploys

CodingLightweight

The featherweight sibling of OpenClaw. ~29k LOC on Bun, docker-per-session, coding-first.

Any Claws model — defaults to claude-sonnet-4-6

64.1k stars · 84k deploys

CodingAutonomous

Nous Research's coding agent. Profiles with per-agent memory, gateways and container-hardened runtimes.

OpenHands

PYTHON · v0.31.0

Any Claws model — defaults to claude-sonnet-4-6

58.4k stars · 92k deploys

AutonomousCoding

Autonomous software-engineering agent that plans, edits code, runs tests and iterates with a web UI.

AutoGen

PYTHON · v0.7.3

Any Claws model — replaces the built-in OpenAI/Azure config

36.1k stars · 71k deploys

Multi-AgentChat

Microsoft's multi-agent conversation framework, with an AutoGen Studio UI for no-code composition.

Letta

PYTHON · v0.9.4

Any Claws model — routed through our unified endpoint

17.2k stars · 41k deploys

MemoryStateful

Stateful, memory-first agents (formerly MemGPT) with a visual Agent Development Environment.

Agno

PYTHON · v2.4.1

Any Claws model — LLM class swapped at container start

26.3k stars · 48k deploys

GeneralPlatform

Full-stack agent platform (formerly Phidata) — batteries-included UI, memory, tools and knowledge.

GPT Researcher

PYTHON · v4.2.0

Any reasoning model — defaults to claude-opus-4-2

15.7k stars · 37k deploys

ResearchReports

Autonomous research agent that produces long-form, cited reports through a clean web UI.

SuperAGI

PYTHON · v0.5.0

Any Claws model — provider config injected at boot

16.9k stars · 29k deploys

AutonomousDashboard

Autonomous long-running agent runtime with a dashboard, marketplace and toolkit ecosystem.