#6666th of 142 in Agent frameworks and SDKsFor science, health and research, robotics

dimos

Python runtime and SDK for controlling robots with agents and MCP skills

Stars
4.7k
License
custom license
Last commit
Nov 2026
Last release
Oct 2026
Language
Python
Stars, last 29 weeks+1,707
21 Apr 2026: 3,035 stars28 Apr 2026: 3,113 stars5 May 2026: 3,147 stars12 May 2026: 3,190 stars19 May 2026: 3,248 stars26 May 2026: 3,307 stars2 Jun 2026: 3,357 stars9 Jun 2026: 3,419 stars16 Jun 2026: 3,469 stars23 Jun 2026: 3,513 stars30 Jun 2026: 3,543 stars7 Jul 2026: 3,612 stars14 Jul 2026: 3,659 stars21 Jul 2026: 3,731 stars28 Jul 2026: 3,784 stars4 Aug 2026: 3,809 stars11 Aug 2026: 3,836 stars18 Aug 2026: 3,870 stars25 Aug 2026: 4,383 stars1 Sept 2026: 4,410 stars8 Sept 2026: 4,446 stars15 Sept 2026: 4,469 stars22 Sept 2026: 4,533 stars29 Sept 2026: 4,571 stars6 Oct 2026: 4,609 stars13 Oct 2026: 4,639 stars20 Oct 2026: 4,664 stars27 Oct 2026: 4,691 stars3 Nov 2026: 4,716 stars10 Nov 2026: 4,742 stars

3,035 on 21 Apr 20264,742 on 10 Nov 2026

From GitHub's daily star counts.

Checked on GitHub.

Overview

DimOS is a Python runtime for robots (quadrupeds, humanoids, arms, drones) built from modules that exchange typed streams and are wired together with composable blueprints. It ships navigation, mapping, perception, spatial memory and manipulation primitives that an LLM agent can call as MCP tools, driven through a `dimos` CLI. Replay and MuJoCo simulation modes run without hardware; a hosted teleop service (dimTELE) is optional.

Who it is for: Robotics developers building agent-controlled robots on Unitree, arms or drones

Strengths

  • Replay and MuJoCo simulation modes run blueprints without robot hardware
  • Hardware list with stability tiers: Unitree Go2, xArm and Piper marked stable
  • Agent skills exposed over MCP and callable from the dimos CLI
  • Transports include Zenoh, LCM, DDS and ROS 2; C++, Lua, TypeScript interop examples

Weaknesses

  • Most hardware support is beta, alpha or experimental; only a few platforms are stable
  • Ubuntu is the CI-tested platform; macOS is alpha and Nix is not CI-tested
  • dimTELE teleop depends on a hosted broker and an API key from Dimensional
  • License is reported as NOASSERTION, so terms need checking in the repo

What it needs

  • Docker + Compose
  • Needs Python 3.12, Git LFS (replay data), MuJoCo (simulation), Ollama (optional, local LLM)
  • Models: LLM agents via MCP, Ollama (local LLM)

What changed

  1. “This release has many breaking changes. Read this section before upgrading.”

  2. v0.0.14 ↗ (opens in a new tab)BreakingSecurity fix

    “This release has many breaking changes. Read this section before upgrading.”

    “Dependabot configured to keep pinned dependencies updated, with a cooldown on newly published Rust crates (supply-chain protection; the matching cargo config takes effect once cargo enforces it). (#1…”

  3. v0.0.13.post1 ↗ (opens in a new tab)BreakingSecurity fix

    “Foxglove viewer support removed (~15k lines); Rerun is the sole viewer backend. (#2122)”

    “Upgrade vulnerable dependencies flagged by Dependabot. (#2040) by @paul-nechifor”

From the release notes. No flag means the note does not say so, not that the upgrade is safe.

Also in Agent frameworks and SDKs

See all 142
Also in Agent frameworks and SDKs
RankProjectScore
1ppt-masterAgent skill that turns documents into natively editable PowerPoint decks Live demo ↗ (opens in a new tab)76.2k stars+15,031 (+25%) in 30 daysMIT96 out of 100
2openai-agents-pythonPython SDK for multi-agent workflows with handoffs, guardrails and tracing30.6k stars+692 (+2.3%) in 30 daysMIT92 out of 100
3mastraTypeScript framework for building AI agents, workflows and MCP servers29.5k stars+778 (+2.7%) in 30 dayscustom license92 out of 100
#4browser-usePython library that lets an LLM agent control a web browser Live demo ↗ (opens in a new tab)120.8k stars+3,334 (+2.8%) in 30 daysMIT90 out of 100
#5copilotkitReact and Angular SDK for adding agent chat and generative UI to apps38.7k stars+757 (+2.0%) in 30 daysMIT90 out of 100