AIエージェント
agenta
A team workspace for building AI agents together.
★ 4.8k スターNOASSERTION
ガイドを見る →invariantは、invariantlabs-ai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: invariantlabs-ai/invariant）。GitHubスターは467、フォークは50で、Apache-2.0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1月12日です。
Guardrails for building secure, robust agents.
Be clear about the problem you are solving before installing.
Clone the repository and skim the docs for hardware or key requirements.
Install dependencies, then add API keys, endpoints, or local-model settings.
Launch the example flow and verify the output matches expectations.
Constrain agent actions and tool calls.
Built to prevent unsafe behavior.
Integrates into the dev workflow.
From a dedicated agent-security team.
Works with Python agent stacks.
Permissive licensing.
Download the source and move into the project folder.
git clone https://github.com/invariantlabs-ai/invariant.git
cd invariantSet up a virtual environment and install the package.
python -m venv .venv && source .venv/bin/activate
pip install -e .Follow the README to guard your first agent.
# see README for the quickstartIt is maintained by Invariant Labs at invariantlabs-ai/invariant, with 50 forks.
Yes. It is released under the Apache-2.0 license, which permits commercial use, modification, and redistribution. Read the LICENSE file for the exact terms.
The library is free; you supply the models and infrastructure.
Compare with Microsoft's governance toolkit on guardrail style and policy language.
Follow the README quickstart to add guardrails to a sample agent.
A team workspace for building AI agents together.
The open-source platform for building and running autonomous AI agents, hosted or self-hosted
Open-source hub of integrations, SDK and CLI for building and deploying GPT/LLM chatbots and agents.
Simple workflow patterns for building effective agents on MCP.
Policy, identity, and sandboxing guardrails for AI agents.
Vercel's open framework for building agents that run anywhere.
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 — invariant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invariant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。
Detect prompt injections and jailbreaks before they hit your LLM.
Snowflake's Agent World Model: synthesise 1,000 executable MCP tool-use environments for agentic RL.
A benchmark testing whether agents can solve long-horizon coding tasks.
A harness-oriented framework for production-grade agent applications.
A graduate-level benchmark even Google can't crack.
The tougher, more robust successor to the MMLU benchmark.