Agent frameworks & infrastructure
Agent frameworks & SDKs
12,596 agents · most starred first
Agent frameworks & SDKs, page 47 of 126
LanchainLLM
Zero-dependency Python helpers to normalize LangChain, MCP, Anthropic content blocks, tool results, and multimodal image payloads.
Phronesis is a Python framework for designing, composing, and operating AI agent systems - typed, observable, and built for production. Agents have episteme (knowledge) from the LLM; Phronesis gives them phronesis (practical wisdom).
An open-source multi-agent AI operating system connecting AI models, agents, MCP servers, and autonomous workflows.
NoWork — open-source desktop AI agent workspace for coding, research, documents, and multi-agent workflows.
Modular .NET 10 platform for software-delivery workflows. Orchestrator + specialist agents (Requirements → Code → Tests → QA) on a provider-agnostic LLM gateway (Claude · Azure OpenAI · MAF · RemoteAgent), multi-tenant runtime with row-level isolation, Blazor "Agent Studio" UI.
Intelligence without consciousness. Process without self. Scale without limit.
Reference architecture for closed-loop AI agents — plan, execute, verify, repair.
Checkpoint/fork/reconcile/rollback SDK for coding agents with dirty-set restoration, crash recovery, branch-conflict detection, benchmarks, and release tooling.
AI-guided onboarding for human-AI collaboration projects
🤖 — a tiny personal AI agent framework. Chat channels (Slack, Telegram, Discord, WhatsApp, …), memory, MCP, and an OpenAI-compatible API in one small Python core.
Specification framework for AI agents and humans
This repo contains demo samples on AI agents using Azure OpenAI, Semantic Kernel, LLM.
Secure, OS-level AI agents built for developers. LangGraph-native framework with built-in guardrails, budget control, and human-in-the-loop.
Self-hosted AI agent that can't say "done" without proof — evidence-gated task contracts, self-distilled skills with auto-rollback, multi-channel.
A unified interface for Large Language Models with memory, reasoning, and tool capabilities. Supports OpenAI, Anthropic, Ollama, HuggingFace, and MLX with hierarchical memory system and agent development features.
A Go compatibility package to allow LLM tool calling for models lacking native tool calling support
Ouroboros is an Agent Harness, CLI and Desktop App with Self-Improvement layer
Partner AI Agent Runtime
Ouroboros, our agent harness: agents with durable memory that live where you talk (terminal, Teams, iMessage, mail, voice) and use real tools. Start with npx ouro.bot@latest.
Typed, auditable tool-use kernel for small LLM agents in Haskell
SPARK Engine — Emotional state, personality, prediction, learning, memory, reflection. The soul layer for autonomous agents.
AI Agent Control Plane — Self-Report, Self-Learn, Self-Evolve. 让 AI Agent 自我管理、自我学习、自我进化。
Python library for building AI agents
Metacognitive self-improving agent framework. TypeScript implementation of DGM-H (Darwin Godel Machine with Hyperagents).
Local-first AI framework: geometric resonance, cognitive architecture, RAG, safety layer, event-driven agentic system. Python 3.13+
Salam dari pulau Batam
The Universal AI Intelligence Layer for Humans and Organizations. CLI with 12+ AI providers, 100+ commands, autonomous agents, automation workflows, and built‑in advisors (finance, legal, business, health).
gradio region:us
gradio region:us
gradio region:us
docker region:us
streamlit region:us
streamlit region:us
gradio agent-demo-track ASE Gemini Gradio region:us
docker region:us
docker region:us
gradio region:us
streamlit region:us
gradio region:us
gradio region:us
transformers safetensors qwen3_5_moe image-text-to-text nvfp4 modelopt
Python framework for personality-steered LLM agents based on Big Five traits, with Ollama integration and research-oriented conversation simulation.
Visual React playground for testing Agentflow agents against a local `agentflow api` server. Launch with `agentflow play`.
Manage multi-turn conversations efficiently with a lightweight C++ library from the llm-cpp toolkit for LLM infrastructure.
Syriana Agent — Open-source AI agent platform. Self-evolving, multi-provider, cross-platform. Works in CLI, Telegram, and Discord. MIT License.
Crash-proof, resumable AI agents — a DeepAgents-style developer experience backed by Temporal durable execution.
AI-native programming language and runtime. Sigma ISA, Omega compiler, CCS cognitive VM. Built for AI agents. Humans and AI as first-class contributors. AGPL-3.0.
Universal Agent Harness
ToolRegistry Hub: Ready-to-Use Tool Collection for LLM Agents — web search, file ops, code execution, and more
Python framework for local-first agent workflows, typed generators, and NFA-based orchestration
Rust library for building tool-using LLM agents: hookable execution, pluggable context optimizers, prompt-cache telemetry, MCP tools, first-class local models (ollama/lmstudio/mlx)
Run an AgentScope 2.0 agent workbench in minutes — no model key required. Native AG-UI + assistant-ui scaffold with SSE streaming, tool calls, and write-tool confirmation.
C++23 framework for building LLM-powered agents with tool use, structured output, multi-provider
NT — a language for agent ecosystems
A tiny testable, self hosted, provider neutral agent runtime for teams that want control, conformance, and no framework lock-ins.
TypeScript SDK for orchestrating Agentic Systems — concurrency, structured output, cost control, and consensus as first-class primitives.
The World's Most Advanced AI Assistant Template — Self-Improving, Multi-Agent, Knowledge-Powered — Built by Friday 🤖
Playing around with LLMs, Langchain, NestJS, Turborepo. Inspired by New Computer's "Dot", but want to make it better.
The thread through any AI labyrinth. C++ LLM workflow orchestration.
Krinik AI Framework (KAIF) — a methodology that brings discipline and autonomy to any AI agent on cognitive projects. A self-extracting, multi-agent system managing the full project lifecycle.
Capability-first vertical agent framework with policy, approvals, workflows, tracing, evals, and an interactive visual lab.
AI Agent Rapid Development Fabric
mupot — installable Cloudflare-native sovereign agent substrate. Fork → deploy to your own CF → log in → run your org (departments, squads, agents). The pot; your business is the plant.
Production-ready CLI AI agent framework with multi-tool orchestration, streaming, sub-agents, MCP integration, lifecycle hooks, configurable safety policies, and a Rich terminal UI.
Production style autonomous agent framework in Python + React that plans multistep tasks, persists state with SQLite, and recovers from failures using exponential backoff retry logic. Includes real time monitoring dashboard, deterministic recovery demo, and API for task orchestration.
Agentic AI Agents - LangChain Agentic Pipeline with Streaming API
Linux of AI is a free, open-source ecosystem for building portable, governed, measurable, and replaceable AI infrastructure. It connects tools for agent orchestration, private deployment, model switching, policy, ontology, audit logs, and AI cost measurement to help reduce vendor lock-in.
Minimal Self-Consciousness Protocol
Turn a use case into a grounded agentic system design. An opinionated AI solutions architect as an agent skill.
Este repositório contém implementações de agentes e ferramentas utilizando a biblioteca LangChain. Ele abrange o uso de Native Tools como PythonREPL, DuckDuckGoSearchRun e WikipediaQueryRun, além de agentes em Python e React. Também explora a integração com bancos de dados SQL usando o SQLDatabaseToolkit.
Realtime multimodal AI agent with voice streaming, RAG memory, and autonomous workflows
A flexible, modularized and easy-to-use agent framework.
Execution-time drift firewall for long-running DeepSeek Harness agents. Real-Harness tests: unsafe stale mutations 12/12 native -> 0/12; valid controls 7/7 both; post-SIGKILL unsafe continuation 2/2 -> 0/2.
A homebrew Agent framework.
Local-first AI Agent framework in Java. ReAct loop, MCP tools, layered memory, multi-agent orchestration — run via Shell, Web or embed it.
Building an Agentic AutoML system from scratch, step by step, from a simple baseline to an experiment-driven senior ML agent.
an agentic core for deploying at scale or building on top of
Open-source Python framework to build, orchestrate and deploy AI agents: multi-provider, tool calling, RAG memory and desktop distribution.
Lightweight browser-side AI agent framework for legacy enterprise web apps—Java 8, Spring Boot 2, TypeScript, MCP/WebMCP, OpenAI-compatible LLMs, tool calling, streaming, audit, and persistence. 面向存量企业 Web 系统的轻量级 Agent 框架。
AutoMind · 本地运行的通用自动化 AI Agent(社区版):分层规划 · 工具调用 · MCP · RAG 知识库 · 飞书/钉钉/企微机器人;Windows / macOS / Linux 桌面版,MIT 开源。
Forge – The AI Execution Layer
Hire AI agents, give them jobs, watch them work. A local-first console where agents own a queue, use real tools, and pass a quality gate before any task counts as done. Zero dependencies, runs fully offline.
a demo plugin for a preset with builtin musescore wasm editor/preview engine for agent composing
A 7-layer architectural blueprint for building reliable, observable, and scalable agentic AI systems in production. Separates concerns like orchestration, context, security, and monitoring to ensure both agent behavior correctness and system performance.
Web dashboard for AgiMate: build agents, set what each one is allowed to touch, and watch the logs
DeepSeek Harness AgentFactory provider for the official Codex App Server
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。
A Clojure-based Agentic AI framework featuring a Unified Concurrent Orchestrator. Executes parallel ReAct loops via local LLMs (llama.cpp) and a polyglot fleet of stateless HTTP microservices (C++, Python, Clojure).
Deterministic, auditable, capability-safe autonomous agent framework in Rust. Event-sourced, replayable, with governed self-evolution.
Friday — a governed agent kernel for Frappe. Agents act as themselves, under their own roles, with every decision audited.
Autonomous self-learning AI agent architecture exploring reinforcement learning, memory persistence, dynamic planning, and tool use.
Evidence-native TypeScript framework for agents that perform bounded work under explicit authority.
DeepSeek Harness 桌面壳 —— 内嵌 Web UI、自动保持最新引擎、自带数据目录管理与系统托盘。
A celld-native AI agent framework — one agent per cell, alarm-paced bounded steps, no platform or model lock-in.
A Rust-based, executable standards system for agentic engineering — featuring versioned standards, substandards, protobuf contracts, automated validation, and agent-native skills.
Python SDK for HelixID — onboard agents, sign and verify verifiable presentations through the API, and delegate authority, with LangChain, CrewAI and MCP middleware integrations. Mirrors the JS SDK's surface.
Embeddable LLM-agent harness for Go — tool-calling, budgets, multi-agent fan-out, stdlib-only core.
A minified agent core: the infrastructure to run an agent, under 2000 readable lines of stdlib Python.
Describe an AI agent, get a ready-to-run agent home with persona, memory and a growth loop. Build one agent or a whole collaborating team.