Skip to content
@strands-labs

strands-labs

Strands Labs

Experimental and cutting-edge projects from the Strands open source community - exploring new frontiers in AI agent development, deployment, and tooling.

Stars Discord

Strands Agents · Documentation · Discord · Discussions


What is Strands Labs?

Strands Labs is the experimental arm of Strands Agents - a space for innovative projects that push the boundaries of what AI agents can do. Labs projects explore new domains, validate novel approaches, and move fast.

What you'll find here

  • New agent capabilities - experimental tools, patterns, and frameworks that extend what agents can do
  • Research-driven projects - ideas validated through real-world usage
  • Community innovation - contributions from teams across the Strands ecosystem exploring uncharted territory

Projects

robots Control robots with natural language through Strands Agents - integrates NVIDIA GR00T and LeRobot
ai-functions Python functions powered by AI agents - with runtime post-conditions for reliable agentic workflows
benchmark-harnesses Strands-based agents and harnesses for agentic benchmarks
harness-optimizer A framework for optimizing LLM agent harnesses through tunable Formulas - dynamically enhancing the agent and improving those Formulas with optimizers based on collected rollout trajectories
pywebrtc-audio Python bindings for WebRTC's audio processing pipeline - noise suppression, echo cancellation, automatic gain control, and voice activity detection via pybind11
strands-for-cosmos NVIDIA Cosmos for Strands Agents. Give your agent eyes that understand physics and hands that generate video, audio, and robot actions - on local compute.
strands-decider A small, fast decision model, or system one model, for agentic workflows. Pick between options or rate on a scale faster than an LLM, with a calibrated confidence on every decision.

Contributing

Have an experimental idea that pushes AI agents forward? Start a discussion or come chat with us on Discord - we'd love to explore it with you.

Popular repositories Loading

  1. strands-decider strands-decider Public

    A small, fast decision model, or system one model, for agentic workflows. Pick between options or rate on a scale faster than an LLM, with a calibrated confidence on every decision.

    Python 1.2k 136

  2. ai-functions ai-functions Public

    Python functions powered by AI agents - with runtime post-conditions for reliable agentic workflows.

    Python 349 43

  3. robots robots Public

    Control robots and physical hardware with natural language through Strands Agents.

    Python 217 41

  4. benchmark-harnesses benchmark-harnesses Public

    Strands-based agents and harnesses for agentic benchmarks.

    Python 72 9

  5. harness-optimizer harness-optimizer Public

    Optimize LLM agent harnesses through tunable Formulas to dynamically enhance the agent, and improve those Formulas with optimizers based on collected agent rollout trajectories.

    Python 68 10

  6. robots-sim robots-sim Public archive

    Simulated environments for robot agent evaluation and reinforcement learning.

    Python 33 10

Repositories

Showing 9 of 9 repositories
  • strands-decider Public

    A small, fast decision model, or system one model, for agentic workflows. Pick between options or rate on a scale faster than an LLM, with a calibrated confidence on every decision.

    strands-labs/strands-decider's past year of commit activity
    Python 1,225 Apache-2.0 136 33 31 Updated Nov 18, 2026
  • robots Public

    Control robots and physical hardware with natural language through Strands Agents.

    strands-labs/robots's past year of commit activity
    Python 217 Apache-2.0 40 0 1 Updated Nov 17, 2026
  • strands-for-cosmos Public

    Give your agent eyes that understand physics and hands that generate video, audio, and robot actions.

    strands-labs/strands-for-cosmos's past year of commit activity
    Python 16 Apache-2.0 7 0 16 Updated Oct 28, 2026
  • ai-functions Public

    Python functions powered by AI agents - with runtime post-conditions for reliable agentic workflows.

    strands-labs/ai-functions's past year of commit activity
    Python 349 Apache-2.0 43 2 5 Updated Oct 21, 2026
  • harness-optimizer Public

    Optimize LLM agent harnesses through tunable Formulas to dynamically enhance the agent, and improve those Formulas with optimizers based on collected agent rollout trajectories.

    strands-labs/harness-optimizer's past year of commit activity
    Python 68 Apache-2.0 10 1 6 Updated Oct 19, 2026
  • .github Public Forked from amzn/.github
    strands-labs/.github's past year of commit activity
    2 566 0 0 Updated Oct 1, 2026
  • benchmark-harnesses Public

    Strands-based agents and harnesses for agentic benchmarks.

    strands-labs/benchmark-harnesses's past year of commit activity
    Python 72 Apache-2.0 9 1 0 Updated Sep 21, 2026
  • pywebrtc-audio Public

    Python bindings for WebRTC's audio processing pipeline, bringing noise suppression, echo cancellation, automatic gain control, and voice activity detection to the Python ecosystem via pybind11.

    strands-labs/pywebrtc-audio's past year of commit activity
    C++ 24 Apache-2.0 8 2 2 Updated Sep 3, 2026
  • robots-sim Public archive

    Simulated environments for robot agent evaluation and reinforcement learning.

    strands-labs/robots-sim's past year of commit activity
    Python 33 Apache-2.0 10 0 0 Updated Aug 6, 2026