AIエージェント
opik
Open-source LLM observability: trace, evaluate and monitor apps and agents
★ 22.5k スターApache-2.0
ガイドを見る →future-agiは、future-agi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: future-agi/future-agi）。GitHubスターは2,127、フォークは671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10月8日です。
One open platform to simulate, evaluate, guardrail, trace, and optimise LLM and agent apps, self-hostable.
Decide whether to self-host (standalone on a laptop or VM, distributed, or Kubernetes) or start with the hosted cloud trial.
Clone the repository and run the installer, which checks your machine, writes local secrets, and downloads the platform images.
Sign in locally, create an account, and copy your API key and secret key from the Keys page.
Instrument your agent with the OpenTelemetry package, send your first trace, then add evaluations, simulations, and guardrails.
OpenTelemetry-native tracing across many agent frameworks, with span graphs, latency, and token-cost dashboards.
Dozens of metrics under one evaluate call, covering areas such as groundedness, hallucination, tool-use correctness, and tone, using LLM-as-judge, heuristic, and ML methods.
Multi-turn conversation simulation against personas, adversarial inputs, and edge cases, for both text and voice agents.
Built-in scanners for issues such as PII and prompt injection, plus adapters for external guardrail vendors, usable inline in the gateway or as an SDK.
An OpenAI-compatible gateway with routing across many providers, semantic caching, and virtual keys.
Several prompt-optimisation algorithms that learn from production traces, closing the loop from monitoring to improvement.
Requires Docker with Compose; the installer prepares secrets and images.
git clone https://github.com/future-agi/future-agi.git
cd future-agi
./bin/installOpen the local app, create your account, and copy the API and secret keys from the Keys page in the sidebar.
# open the installed web app locally and copy FI_API_KEY / FI_SECRET_KEY from KeysInstall the instrumentation package, export your keys, and emit a span from Python.
pip install fi-instrumentation-otel
export FI_API_KEY=<your key>It is maintained by future-agi on GitHub, alongside a commercial cloud offering.
The platform core is Apache-2.0 licensed and free to self-host; the company also offers a managed cloud and enterprise editions.
Not for the platform itself, but evaluations and optimisation call LLMs, so you configure your own provider keys for those features.
Tools such as Langfuse or Phoenix focus mainly on tracing and evals; Future AGI bundles simulation, guardrails, a gateway, and prompt optimisation into the same platform, per its own comparison.
Clone the repo, run ./bin/install, sign in locally, copy your keys, and pip install fi-instrumentation-otel to send a first trace.
Open-source LLM observability: trace, evaluate and monitor apps and agents
A low-code tool for building multi-agent LLM apps — chatbots, RAG and fine-tuning in one flow.
iFLYTEK's enterprise agentic workflow platform — SuperAgents with workflows, RPA and one-click deploy.
Make humans and AI agents work as one team — open-source and self-hostable.
One portable, local-first memory layer for every AI agent - Markdown-native, user-owned and self-evolving.
Code, build and evaluate LLM agents with full MCP, Skills, ACP and A2A support.
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 — future-agi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future-agi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。
Shrimp Task Manager: an MCP server that turns natural language into structured, dependent dev tasks.
Open-source AI assistant framework with skills and an OpenClaw-style agent core.
A benchmark for tool-agent-user interaction in real-world domains.
Uber's ADR: discover, observe, benchmark, and detect threats across enterprise AI agents, deployed at Uber.
Open Cowork: a desktop AI agent app with one-click Claude Code, MCP tools, skills, and VM sandboxing.
Manage parallel AI coding agents on an infinite tmux-backed canvas.