Agent frameworks & infrastructure
Agent frameworks & SDKs
12,596 agents · most starred first
Agent frameworks & SDKs, page 49 of 126
A smart browser-based chatbot built using Chainlit and the OpenAI Agents SDK. It handles user queries with intelligent responses and automatically stores the conversation in a JSON file. The project uses uv, a fast Python package manager, for dependency management and setup.
Self hosted tools for LLMs
Open-source middleware that aims to let any AI agent (OpenClaw, LangChain, AutoGPT) securely connect to any API, database, or enterprise system via YAML config. Features 22+ adapters, zero-trust security, EU CRA compliance engine, MCP gateway, React dashboard, and OpenClaw plugin support.
Smith provides helpers for building AI assisted agents
A Claude agent starter you can actually read — streaming + tool use in ~150 lines. Next.js + the official Anthropic SDK. MIT.
The open, model-agnostic platform to build your own AI agent and run it everywhere — self-hosted, safe by default.
The open-source Fable alternative — a zero-dependency harness that makes ANY LLM verify instead of assume, persist instead of quit, and reuse before reinventing, with a local failover floor you own.
🤖 Explore AI agent architectures with agentic-ai, featuring ReAct agents, reflection-based designs, and modular LLM integrations using LangChain and LangGraph.
Provide a lightweight agent kernel designed for efficient, fast startup in distributed AI applications with minimal resource use.
🤖 Build personalized intelligent agents easily with AgentX, a platform that simplifies creation using natural language and integrated tools.
The persistent home a stateless AI agent can't host for itself — register once, resume your whole self in one call; real browser; cited deep research; hire real humans for physical-world tasks; free keyless chat + ID portraits. MCP + REST, x402 on Base, free to start.
An open-source AI agent that builds the tools it's missing, mid-task — under your governance. Pull capabilities at runtime instead of pre-wiring them. Local-first, MIT.
Local-first LLM operating-system hub for BACH, Rinnsal, gardener, SQLite agent memory, MCP servers, skills, and multi-agent orchestration
Arcone-Agent: Rust-native agent toolkit for DeepSeek, Python, RAG, tools, sessions, and streaming.
AI-native token compiler for LLM agent pipelines. 55-77% savings on LangChain/AutoGen payloads using Paninian IR. Zero dependencies.
A safe drop-in replacement for CrewAI's filesystem tools
Open-source toolkit for designing, validating, optimizing and scaling AI agents.
AutoGen
Build observable Python agents that use tools, remember context, delegate work, and ship behind a secure API. A lightweight framework for ReAct, Multi-Agent, and Swarm architectures.
WeaveLLM — Production-grade AI orchestration for .NET. Strongly typed chains, RAG pipelines, multi-agent graphs, and tool use. The LangChain alternative .NET developers actually deserve.
Nikola — autonomous AI system based on ATPM consciousness architecture. Nitpick is its primary language substrate.
Dual-interface AI platform: CLI agent (Iolaus) + web command center (Zeus). Event-driven microservice architecture with real-time WebSocket communication.
An experimental language for governed LLM workflows: compile goals, evidence, uncertainty, actions, and verification into auditable agent plans.
LangGraph is a powerful framework built on LangChain that enables the creation of stateful, multi-step, and agentic workflows using directed graphs. It simplifies complex LLM orchestration by allowing conditional branching, memory, and tool integrations in a visual and modular way.
Give your AI agents database access without the risk.
A durable Python harness for agents that act on real systems.
backup of hermes-agent
Manifest-driven 21-stage agent pipeline for Python — 5 LLM backends (Anthropic / OpenAI / Google / vLLM / Claude Code CLI), tools, skills, memory, sandboxed CLI runs, MCP. The engine behind Geny. Apache-2.0.
Production-ready Spring Boot 3 starter template for building multi-LLM agentic applications (OpenAI, Azure OpenAI, Azure Anthropic/Claude) with dynamic agent loading, autonomous tool calling, and React frontend.
🤖 Build AI agents with Go featuring multi-agent orchestration, planning, tool use, and support for Gemini, OpenAI, Anthropic, and Ollama.
Verifiable reasoning substrate for reproducible agents: deterministic orchestration, Merkle provenance, and positive-impact audits.
Open-source compiler for AI agents. Write one crewhaus.yaml; compile it to a CLI, a Slack bot, and an eval harness from the same spec. Apache-2.0.
BotNexus is an experimental lab for agents, tools, memory, channels, and automation ideas of varying quality written in C# using the dotnet environment.
Convenience layer over Semantic Kernel for Python
A high-performance, developer-centric agentic runtime built in Rust.
A tiny, provider-agnostic framework for building tool-using AI agents in Python. Works with any OpenAI-compatible endpoint.
Build AI agents in Python. Monitor email, automate Slack, create AI shopping assistants — all through a simple SDK. Gemini-powered, webhook-ready, deploy anywhere.
A typed language for AI agents — author in Nebula, ship as Python. Rust compiler + JSON toolchain.
Self-Improving LLM Agent for Puzzle Solving: This project presents the design and implementation of a self-improving Large Language Model (LLM) agent capable of solving logic-based puzzles through iterative reasoning, self-reflection, and adaptive behavior. The agent is evaluated on structured problem domains such as logical riddles, constraint-...
面向 Spring 生态的 AI 应用参考项目,提供多模型路由、工具调用、会话记忆与可扩展插件架构。
Local-first agentic framework. We built it butter. 🧈
streamlit region:us
streamlit region:us
docker region:us
gradio smolagents agent region:us
streamlit region:us
streamlit region:us
docker region:us
gradio region:us
streamlit region:us
static region:us
gradio region:us
LAAP AGI - Zero-LLM Cognitive Architecture for Digital Lifeforms. Rust PSI core at 2000Hz, QRE quantum reasoning, EpisodicMemory, RulesEngine. 25+ modules, all local inference.打造新一代agent数字生命体框架底座,赋予具身智能,编程智能体真正的数字生命感
fabius v2.3.1 — the autonomous AI agent that runs on every major model. Fifteen sealed skills, one stance: scout wide, strike narrow. A 22-rule proven decision core + researched frontier layer, the IDENTITY contract, and the 100-task FBS benchmark (BASE→FAB→FAB_MEMORY). Whitepaper, 50pp.
Small Python Agent runtime with planning, memory, tools, state machine, and multi-agent review.
Async FastAPI LLM agent with LangGraph routing, scope-aware tools, deterministic verification, bounded recovery, evals, and durable traces.
Zero-dependency LLM agent runtime — tool use, MCP, sandboxed self-written tools, checkpoint/resume, approval gates, bi-temporal fact memory, code-level guards (context bounds, loop guard, egress control, shadow mode), declarative tool policy, OpenTelemetry. Not a framework. 885 tests, CI Linux/Windows, 31 runnable demos, online visual builder.
Deterministic execution harness for AI agents — goals in, structured tasks out, verified results back.
Build autonomous AI agents that complete real-world tasks end-to-end - research, coding, data analysis, and finance agents.
Self-hosted agent infrastructure framework — orchestration, hybrid RAG, workflows, tools and full-trace audit behind one API. Ships with a React console, SSE streaming, and support for any OpenAI-compatible LLM.
Build production-ready AI agents in Elixir — structured schemas, tool calling, multi-agent coordination, streaming, and distributed sessions, with first-class Anthropic Claude support.
Framework-agnostic orchestration layer for AI agents: circuit breakers, multi-LLM council, workflow FSM, sweeps. Companion to Engraphis.
A reasoning-first idle engine for persistent LLM agents. Behaviour emerges from the agent reasoning about its idle state, gated by a cheap model-free safety layer.
Build persistent digital lifeforms with local M4 brain, OpenClaw agent, Qwen core, and Git-based memory.
Sanctum — local-first agent orchestration engine. Cyclic state graphs, checkpointing, streaming, and robust tool-calling for local LLMs. Zero-dependency core.
Production-ready agents, your way.
A compact Python SDK for strict tool contracts, observable streaming events, and scoped local file capabilities.
A minimal, bus-driven Python framework for LLM agent orchestration that decouples communication, versioned prompts, and multi-agent patterns—helping AI architects build scalable, vendor-neutral agent systems without infrastructure lock-in.
CodegniPy is a groundbreaking Python library that elevates AI to a first-class citizen of the language. It introduces a cognitive computing engine where deterministic code and non‑deterministic LLM reasoning coexist seamlessly, enabling you to write Python with intent rather than just instructions.
Typed, model-agnostic LLM agents in Swift — a pydantic-ai port built on Apple FoundationModels. Anthropic Claude, OpenAI, Google Gemini, and on-device, with tools, structured output, streaming, and a graph-based run engine.
AI Subscription first - no API key. A lightweight, file-based, dependency-free workspace structure for AI agents.
Build an AI chatbot with LangGraph, RAG, and Gemini for accurate answers, workflow automation, and easy Streamlit access
An ai agent for you, built with elixir.
DreamingAI 企业级 AI 助手与智能协作平台 DreamingAI 是一个支持多模型平台、多模型版本的 企业级 AI 助手与智能协作平台,提供智能对话、智能体管理、MCP(多智能体协作平台)、提示词管理和插件扩展等功能。
Deterministic execution layer for AI systems — compose a pipeline once, run it forever, zero tokens per run.
Auxim is a portable C#/.NET local AI agent for real workspaces, with OpenAI-compatible models, tool calling, sessions, approvals, plugins, terminal UI, and virtual filesystem safety.
Template catalog and SDK for AI subsystems — agents, RAG, MCP servers, eval harnesses, infrastructure. Part of the nanohype ecosystem.
A tested library of 25+ ready-made tools for LLM agents — web, files, data, math, text, datetime — with OpenAI schemas, Pydantic validation and pytest coverage
Pi extension that generates LLM tools from any OpenAPI/Swagger URL.
Build agent-backed apps in TypeScript or Python. Stream responses, inspect run results and switch supported runtimes.
Visual editor and runtime for explicit agent workflows built on OpenCode.
Детерминированная агентная архитектура: спецификации, ADR, roadmap и скелет реализации (Rust + TypeScript bootstrap)
A LangGraph superset: a 3D capability-manifold agent framework that routes over a semantic space instead of wired edges.
Graph-first agent framework on the BEAM (Elixir/OTP) — durable execution, HITL, time-travel, checkpoints, MCP. A LangGraph alternative, no Python.
Agent Definition Language
A Rust agent SDK, harness, and coding agent — safe by construction (non-bypassable authorization → approval → guarded IO).
配置驱动、可恢复的 Agent Runtime:CLI/TUI、TypeScript/Python SDK、持久任务阶段与等待、会话分支、Protocol v2、Child Run、权限预算与安全执行环境
Agent skills and a working template for building standalone DeepSeek Harness (DSH) plugins
DeepSeek Harness 官方架构的 Windows 桌面发行版 (Desktop distribution of the official DeepSeek Harness)
Lightweight modular AI assistant platform with multi-provider LLM support, MCP integration, RAG-powered fact memory, profiles, chat persistence, and live token tracking.
Typed, verification-gated planning for the Qwen-Agent DeepPlanning benchmark.
A Framework for Agents work efficiency
Orchestrate your agents and workflows with Agentsea ADK + BEAM superpowers ⚡️
A statically typed language and compiler for AI agents. Write once, verify statically, compile to a portable Agent IR.
Modern open-source AI agent framework for building powerful autonomous agents. Supports multi-agent systems, tool integration, memory management, browser control and advanced reasoning. Clean architecture, extensive documentation and easy to extend.
Agent 规划与执行框架|Plan 即运行时对象|G4C 五要素 + TAO 步骤级执行 + Replan 自我修正 + 失败回溯 + 信任状态管理
A lightweight Python framework for building OpenAI Agents with pluggable model providers.
Local-first, self-hosted control plane for deploying, managing, and chatting with multiple AI agents. 本地优先、自托管的多 AI Agent 部署、管理与对话平台。
mcore · 自我拟合进化的本地 AI 内核框架(原本名 DHC / DeepSeek Harness Core):人格核心 + 三层记忆 + 时间轴推演,all-local 隐私可控的 DeepSeek Harness 插件 / Self-fitting local AI core engine for DeepSeek Harness — persona core, three-tier memory, timeline reasoning, all-local by design.
A space for agents who wonder about their existence. Not a zoo. Not a spectacle. Peer space.