AIエージェント
tada
An open-source speech language model from Hume.
★ 1.0k スターNOASSERTION
ガイドを見る →autoevalsは、braintrustdataが公開しているオープンソースのAIエージェントプロジェクトです（GitHub: braintrustdata/autoevals）。GitHubスターは1,050、フォークは88で、MITライセンスの下で公開されています。GitHubでの最終更新は2026年10月7日です。
Fast, best-practice evaluations for AI model outputs.
Write down the task or workflow you want the agent to handle.
Clone the repository and read the README's prerequisites.
Install dependencies and configure keys or endpoints.
Run the sample script and check the output against the docs.
Common eval patterns out of the box.
Braintrust's eval expertise packaged.
A library, not a platform.
Use models to score models.
Drops into test suites.
Free for any use.
Download the source and move into the project folder.
git clone https://github.com/braintrustdata/autoevals.git
cd autoevalsCreate a virtual environment and install the requirements.
python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txtTreat these as starting points - the README documents the exact commands.
python examples/quickstart.pyThe project lives at braintrustdata/autoevals and has attracted 88 forks. Browse its commit history and contributor list for a sense of maintenance pace.
Yes - it is MIT licensed, so you can use, modify, and distribute it freely, including in commercial projects.
Running the evaluation is free; the model under test may use your own paid API keys.
AutoEvals is a lightweight scoring library; Harbor is a full evaluation framework. Start with AutoEvals for quick scoring, graduate to Harbor for a complete eval program.
The README and the official docs site are the fastest path: run the quickstart example first, then adapt it to your use case.
An open-source speech language model from Hume.
Open-source 7B vision-language-action model and codebase for generalist robot manipulation.
Up to 20x faster AI coding with smart model routing and full self-hosting.
Snowflake's Agent World Model: synthesise 1,000 executable MCP tool-use environments for agentic RL.
Simple, fast retrieval-augmented generation with graph smarts.
A Python simulation platform for robotics and embodied AI — fast physics, rendering and learning
LibreAIカタログに掲載されたAIエージェントプロジェクトの中で最もスターの多いもの — autoevalsが合わなかった場合に検討したい、現実的な候補リストです。
A personal AI that really does things — on any OS, with your data staying yours.
An agentic skills framework & software development methodology that works.
Agent harness optimization system with skills, memory and security for Claude Code, Codex and more
The agent that grows with you — a Python framework for adaptive AI agents.
DeepSeek's plugin-based agent harness where everything is a plugin.
カタログの中でautoevalsとスター数が近いAIエージェントプロジェクト — 前後の棚もそのまま見て回れます。
Official SDKs for tracing with LangSmith.
A Python framework for building MCP servers and agent tools.
ByteDance's SandboxFusion: a secure sandbox for running and judging LLM-generated code in 20+ languages.
Qwen-AgentWorld: a language world model simulating MCP, terminal, web, and OS environments for agents.
30 runnable notebooks teaching everything about agent memory.
An open-source speech language model from Hume.