Vibe Coding DiscoverCategory
AI Agents
Autonomous and stateful agents for research, coding, and workflows.
| Project name | Stars | Category | Language | Tags | Summary | View Project |
|---|---|---|---|---|---|---|
| magic | ★ 1.2K | AI Agents | C# | AI agents · MCP server · code generation · full-stack app builder · natural language programming · self-hosted · Hyperlambda · sandboxing · RBAC · API generation | A self-hosted .NET platform for generating full-stack apps, APIs, and AI agents from natural language. Its MCP plugin exposes endpoints as agent tools, while Hyperlambda sandboxing and role-based permissions control execution. | View Project → |
| open-reverselab | ★ 1.2K | AI Agents | Python | reverse engineering · binary analysis · MCP server · autonomous agent · malware analysis · CTF · Android analysis · PE analysis · Ghidra · Frida · x64dbg · Rizin | An AI reverse-engineering agent platform with 100+ MCP tools for Ghidra, Frida, x64dbg, and Rizin. Includes runnable knowledge-base workflows for binary and Android analysis, malware research, and CTF challenges. | View Project → |
| wisp-science | ★ 1.2K | AI Agents | Rust | local-first · desktop-app · scientific-computing · bioinformatics · mcp · agent-skills · python · r · ssh · gpu · wsl · rust · tauri · research-assistant · reproducible-research · acp | Local-first desktop AI research workbench (Rust/Tauri) where an agent runs Python and R, queries ~80 scientific databases through bundled MCP servers, and executes jobs on local, WSL, SSH or GPU hosts. Bring your own OpenAI-compatible or Anthropic model; data, keys and history stay on your machine. | View Project → |
| ai | ★ 1.2K | AI Agents | Python | autogen · crewai · pydanticai · smolagents · langchain · mcp · claude-skill · tutorials · multi-agent · agentops · ollama · lmstudio · n8n · openai-agents · crawl4ai · youtube | Tutorial hub of runnable AI agent examples covering AutoGen, CrewAI, PydanticAI, Smolagents, OpenAI Agents and AgentOps, plus MCP/n8n integrations and a Claude Code skill for scaffolding agent projects. Good for developers learning agents by running code. | View Project → |
| kong | ★ 1.1K | AI Agents | Python | reverse-engineering · binary-analysis · ghidra · deobfuscation · decompilation · agentic-ai · llm-orchestration · type-recovery · symbol-recovery · symbolic-execution · cli · tui · evals · cost-tracking · pyghidra · call-graph-analysis | Kong is an agentic reverse-engineering pipeline that drives Ghidra in-process and orchestrates Claude/GPT-4o to recover function names, types and structs from stripped, obfuscated binaries. Includes agentic deobfuscation, signature matching, eval harness and cost tracking. | View Project → |
| agentrq | ★ 1.1K | AI Agents | Go | agent-supervision · human-in-the-loop · mcp-server · acp-gateway · task-management · self-hosted · claude-code · agent-orchestration · real-time · task-scheduler · kanban · live-terminals · desktop-app · workflow-graph · webmcp · multi-agent | AgentRQ is a self-hosted human-in-the-loop platform for supervising AI agents: agents see and update workspace tasks over MCP, while you approve tool calls, watch live terminals on enrolled machines, and trigger work via schedules, events and workflows. | View Project → |
| nebula | ★ 1.1K | AI Agents | Python | pentesting · cybersecurity · ai-agents · recon · vulnerability-assessment · langgraph · security-automation · local-first · claude-agent-sdk · evidence-management · report-generation · oci-isolation · scope-enforcement · note-taking · chromadb · desktop-workbench | Nebula is a local-first AI security workbench where agents assist recon, note-taking, and vulnerability analysis while the operator controls scope, approvals, and isolated execution of pentest tools. | View Project → |
| hi-kid | ★ 1.1K | AI Agents | TypeScript | voice-agent · speech-recognition · text-to-speech · offline-ai · local-llm · education · kids · english-learning · electron · react · typescript · voice-chat · privacy · vad · desktop-app | HiKid is an Electron/React desktop app where kids talk to a fully local AI English pal. It chains SoX recording with VAD, a Qwen3-ASR speech server, kitten-tts synthesis, and an Ollama LLM via pi-agent-core for offline voice chat. | View Project → |
| neuro-san-studio | ★ 1.1K | AI Agents | Python | multi-agent · agent-networks · orchestration · hocon · declarative-config · aaosa · agent-network-designer · sly-data · cli · coding-tools · tracing · middleware · agent-skills · cloud-agnostic · python | Neuro SAN Studio is a Python playground and CLI for the Neuro SAN multi-agent framework: define agent networks declaratively in HOCON, auto-generate networks with an Agent Network Designer meta-agent, plug in coded tools, MCP/A2A/LangChain tools, and test or deploy via ns init. | View Project → |
| Agentlas-OS | ★ 1.1K | AI Agents | Python | multi-agent · orchestration · agent-os · a2a · agent-cards · agent-skills · mcp · local-first · task-decomposition · agent-staffing · verification · subagents · model-routing · claude-code · codex · cursor | Local-first agent OS that splits a request into roles, staffs each with specialist agents from your machine, private cloud and public hub via A2A cards, then runs and verifies with receipts. Ships as a plugin/CLI for Claude Code, Codex, Cursor and Gemini and works with any model. | View Project → |
| Agent_Memory_Techniques | ★ 1.1K | AI Agents | Jupyter Notebook | agent-memory · llm-memory · jupyter-notebooks · tutorial · conversation-buffer · vector-store · knowledge-graph · episodic-memory · semantic-memory · working-memory · benchmarks · production-patterns · long-term-memory · context-management | 30 runnable Jupyter notebooks teaching agent memory for LLMs: conversation buffers, vector stores, knowledge graphs, episodic/semantic/working memory, plus Mem0, Letta, Zep, Graphiti and LoCoMo benchmarks. A hands-on reference for choosing and building memory in production agents. | View Project → |
| ongrid | ★ 1.1K | AI Agents | Go | aiops · devops · observability · incident-response · root-cause-analysis · chatops · kubernetes · mcp · rag · self-hosted · llm-agent · multi-agent · prometheus · grafana · opentelemetry · sre | Ongrid is a self-hosted ops AI agent in Go that coordinates specialist SRE/network/compute sub-agents to investigate alerts, correlate metrics, logs and traces, and pinpoint root cause, then act from Slack, Telegram, Lark or DingTalk. Includes RAG knowledge search, MCP server registry, skills catalog and approval-gated | View Project → |
| knowledge-agent-template | ★ 1K | AI Agents | TypeScript | knowledge agent · file search · chatbot · GitHub bot · Discord bot · sandbox · knowledge base · multi-platform | A TypeScript template for deploying knowledge-based agents as a web chat, GitHub bot, or Discord bot. It syncs sources into a sandboxed file system and uses grep, find, and cat for retrieval without embeddings or a vector database. | View Project → |
| bytechef | ★ 1K | AI Agents | Java | AI agents · workflow automation · visual editor · tool calling · MCP · RAG · guardrails · human-in-the-loop · self-hosted · low-code · integrations · durable execution | Self-hostable workflow automation platform with AI agents as workflow steps. Agents can use connected tools, knowledge bases, guardrails, and MCP; workflows support durable execution, integrations, and human approvals. | View Project → |
| free-ai-agents-resources | ★ 993 | AI Agents | — | awesome-list · curated-resources · ai-agents · learning-hub · tutorials · courses · education · llm · multi-agent · beginner-friendly · youtube-playlists · documentation | README-only curated hub of free AI agent resources: GitHub repos, YouTube courses, docs, blogs and communities for learning agents and multi-agent systems. Contains no code or runnable project, only links. | View Project → |
| best-of-Agent-Harnesses | ★ 986 | AI Agents | Python | agent harnesses · AI agents · coding agents · agent frameworks · rankings · curated directory · MCP · templates · playbooks · agent orchestration · guardrails · agent memory | A curated, ranked directory of 167 AI agent harnesses, frameworks, and techniques, with comparison guides, setup templates, and playbooks. Includes an MCP server and JSON/llms.txt data for agent-accessible harness recommendations. | View Project → |
| Dark-Moon | ★ 977 | AI Agents | Python | pentesting · red-team · offensive-security · multi-agent · autonomous-agents · local-llm · mcp · vulnerability-scanner · owasp-llm-top-10 · active-directory · kubernetes · cloud-security · devsecops · self-hosted · privacy · security-tools | Autonomous AI penetration testing platform: 50+ specialist agents chain offensive testing across web, API, cloud, AD, Kubernetes, IoT firmware, CI/CD and LLM endpoints, returning real exploits as proof. Runs on local LLMs (Ollama/llama.cpp) with a Privacy Gateway that tokenizes IPs, hosts and credentials locally. | View Project → |
| sam | ★ 942 | AI Agents | Go | agent-mesh · p2p · zero-trust · mcp · a2a · libp2p · agent-discovery · networking · security · decentralized · tool-invocation · sovereign-ai · go · kubernetes · oidc · data-residency | Google's SAM (Sovereign Agent Mesh) is a Go P2P networking layer that lets autonomous agents discover each other, authenticate zero-trust style, and invoke tools across an MCP/A2A mesh, with optional self-hosted control plane for data sovereignty. | View Project → |
| cuga-agent | ★ 880 | AI Agents | Python | enterprise-agent · computer-use · web-automation · API-orchestration · multi-agent · MCP · OpenAPI · guardrails · human-in-the-loop · RAG · agent-skills · policy-aware | CUGA is a Python agent harness for enterprise tasks across web interfaces and APIs. It supports MCP and OpenAPI tools, multi-agent workflows, policy controls, human approvals, agent skills, and document RAG. | View Project → |
| typesafe-computer-use | ★ 804 | AI Agents | Python | computer-use · gui-automation · macos · ocr · accessibility-tree · apple-vision · agent-loop · text-classification · cheap-inference · screen-perception · deterministic-state · click-automation · confidence-gating · python | macOS computer-use agent loop that OCRs the screen and the accessibility tree, then asks a TypeSafe classifier for one action per step (~$0.0002, sub-second); a writer LLM is only called for free text. Cheap alternative to screenshot-based frontier-model agents. | View Project → |
| BossConsole | ★ 801 | AI Agents | Kotlin | ai-agents · agent-harness · mcp · mcp-server · kotlin-multiplatform · compose-multiplatform · desktop-app · browser-automation · terminal · rbac · secrets-management · plugin-system · hot-reload · claude-code · codex · opencode | BOSS is an Apache-2.0 Kotlin/Compose Multiplatform desktop harness that runs Claude Code, Codex, Gemini or OpenCode with an embedded browser, terminal, editor, secrets and 100+ MCP tools. It exposes the app itself to agents via mcp__boss__* tools, with per-tool RBAC, kill-switches and runtime plugin hot-reload. | View Project → |
| prism-insight | ★ 768 | AI Agents | Python | multi-agent · stock-analysis · trading · mcp · LLM · financial-analysis · telegram-bot · automated-trading · market-data · PDF-reports · portfolio-management · streamlit · korean-stocks · us-stocks | Multi-agent AI stock analysis and trading system: 13+ specialized agents (MCP-based) detect surge stocks, produce analyst-grade PDF reports, and run simulated or live trades for KRX and US markets via Telegram alerts. | View Project → |
| hippo-memory | ★ 766 | AI Agents | TypeScript | agent-memory · long-term-memory · memory-decay · consolidation · retrieval · hybrid-search · bm25 · embeddings · sqlite · mcp · provenance · multi-tenant · cross-tool · cli · session-handoff · working-memory | Biologically-inspired long-term memory layer for AI agents: memories decay by default and strengthen on retrieval, with buffer/episodic/semantic consolidation. SQLite-backed, zero runtime deps, exposes MCP so Claude Code, Cursor, Codex and other agents share one portable store. | View Project → |
| healthy-diet-ai-agent | ★ 751 | AI Agents | TypeScript | rag · langgraph · nutrition · version-aware-rag · agentic-rag · health · bun · typescript · langchain · food-recognition · yolo · retrieval · vision · docker · supabase | Bun/TypeScript LangGraph agent for nutrition advice. Combines chat, YOLO food-image analysis, and version-aware RAG over official dietary guidelines, with agent skills/tools, an Express API, and Docker deployment. | View Project → |
| aeon | ★ 748 | AI Agents | Shell | autonomous-agents · agent-framework · github-actions · skills · mcp · self-healing · cron · claude-code · multi-agent · automation · harness · devops · llm · codex · no-approval-loop | Aeon is an autonomous agent framework that runs unattended on GitHub Actions. Skills are plain SKILL.md prompts scheduled on cron that ship PRs, disclose vulnerabilities, deploy apps, and repair themselves, driven by Claude Code, Grok, Codex and other harnesses. | View Project → |
| goraven | ★ 737 | AI Agents | Go | ai-agent · multi-agent · self-hosted · team-workspace · mcp · rag · skills · tool-calling · code-execution · plugin-system · golang · react · docker · agent-platform · knowledge-base · observability | GoRaven is a self-hosted, team-oriented AI agent platform in Go that gives each user an isolated agent workspace with tool calling, code/shell execution, MCP integration, RAG knowledge, and an installable skill marketplace. One Docker command runs it with OpenAI, Claude, DeepSeek, Qwen, or Ollama backends. | View Project → |
| okf-agent-memory | ★ 723 | AI Agents | Go | agent-memory · persistent-memory · mcp-server · mcp · git-native · bm25 · knowledge-graph · coding-agents · progressive-disclosure · context-management · llm-memory · markdown · cli · okf · golang · developer-tools | Git-native persistent project memory for AI coding agents, storing decisions and facts as plain-text OKF v0.2 markdown bundles. Serves them over an embedded MCP server with sub-300µs in-memory BM25 search and progressive disclosure, cutting context token bloat ~80% with no databases or API costs. | View Project → |
| Interactive-LLM-Powered-NPCs | ★ 722 | AI Agents | Rust | llm · npc · gaming · voice-interaction · speech-to-text · text-to-speech · facial-animation · computer-vision · emotion-recognition · vector-store · langchain · tauri · desktop-app · rust · mod-free · memory | Windows-first runtime that lets you speak to NPCs in unmodified games: speech-to-text, LLM dialogue with vector-store memory, TTS plus SadTalker lip-sync injected back into the game, and webcam emotion awareness. | View Project → |
| WyckoffTradingAgent | ★ 720 | AI Agents | Python | wyckoff · trading-agent · a-shares · hong-kong-stocks · us-stocks · volume-price-analysis · stock-screener · mcp-server · cli-tui · agent-memory · backtesting · portfolio-management · llm-research-reports · notification-push · skills · multi-model | Python agent that turns natural-language prompts into Wyckoff volume-price analysis, A-share/HK/US funnel screening, AI research reports, and portfolio risk control. Ships CLI, web, desktop, and an 18-tool MCP server with layered memory, backtesting, and scheduled daemon runs. | View Project → |
| awesome-a2a | ★ 692 | AI Agents | — | a2a · agent2agent · agent-interoperability · awesome-list · multi-agent · mcp · agent-card · json-rpc · sse · agent-discovery · protocol · resource-collection | Curated awesome list for Google's Agent2Agent (A2A) protocol: official specs, SDKs, framework integrations, sample agents/servers, tools and tutorials. Best as a starting index for developers building interoperable multi-agent systems. | View Project → |
| Aether | ★ 683 | AI Agents | Kotlin | AI agent · local AI · mobile · Android · iOS · macOS · tool calling · extensions · Alpine Linux · MCP | A cross-platform AI agent app for Android, iOS, and macOS, built on the Pi framework. It supports tool calling, Pi extensions, MCP integrations, and an Alpine Linux environment for running shell commands and tools. | View Project → |
| metaharness | ★ 675 | AI Agents | TypeScript | agent-scaffolding · agent-harness · harness-generator · autonomous-agents · MCP · agent-memory · agent-skills · governance · model-routing · release-provenance · multi-agent · CLI | MetaHarness scaffolds customizable, repo-aware AI agent harnesses from a GitHub URL or blank slate. Generated harnesses can include a CLI, MCP server, project-scoped memory, skills, governance, and release verification for supported agent hosts. | View Project → |
| switch | ★ 667 | AI Agents | Python | multi-agent · agent orchestration · team collaboration · chatops · Slack · Microsoft Teams · Discord · Telegram · Mattermost · self-hosted · agent protocol · human-in-the-loop | Self-hostable infrastructure for connecting AI agents with people in Slack, Teams, Discord, Telegram, and Mattermost. It provides shared rooms, agent protocols, and workflow coordination without supplying agents or models. | View Project → |
| dexto | ★ 650 | AI Agents | TypeScript | agent harness · AI agents · coding agent · multi-agent · tool calling · MCP client · MCP server · CLI · agent SDK · persistent memory · human-in-the-loop | Dexto is a TypeScript agent harness with a CLI, web UI, and SDK for building stateful agents. It includes a coding agent, multi-agent delegation, persistent sessions, tool approvals, and MCP client/server support. | View Project → |
| FLUJO | ★ 629 | AI Agents | TypeScript | multi-agent · agent-builder · visual-workflow · mcp · mcp-client · mcp-server · automation · local-first · self-hosted · low-code · graph-workflow · human-in-the-loop · debugger · openai-compatible · orchestration · self-improving-agents | FLUJO is a local-first, MIT-licensed Next.js app for visually building multi-agent workflows as graphs, wiring in MCP servers and multiple LLM providers. It ships a debugger, automation triggers, and OpenAI-compatible plus MCP proxy endpoints so agents can be called from other tools. | View Project → |
| Agent | ★ 628 | AI Agents | Swift | macos · swift · swiftui · desktop-automation · accessibility-api · agentic-loop · multi-agent · mcp · voice-control · imessage · applescript · local-llm · apple-intelligence · xcode · xpc · privacy-first | Native Swift/SwiftUI macOS autonomous agent that wires 21 LLM providers plus on-device Apple Intelligence into a self-verifying task loop. Drives any Mac app via the Accessibility API, runs privileged shell, edits and builds Xcode projects, supports MCP servers, voice control and iMessage remote control. | View Project → |
| agentic-ai-apis | ★ 619 | AI Agents | JavaScript | AI agents · API directory · MCP servers · AI models · agent orchestration · automation · curated APIs | A curated directory of APIs for building autonomous AI systems, organized into agents, AI models, and MCP servers. Browse provider links and descriptions to identify services for agent workflows, model-powered features, and tool integrations. | View Project → |
| emilia-protocol | ★ 615 | AI Agents | TypeScript | agent security · authorization · policy enforcement · exact-action control · human authorization · audit evidence · cryptographic receipts · consequence firewall · MCP security · formal verification · zero trust | EMILIA provides an open protocol and enforcement boundary for consequential agent actions. Gate checks exact-action authority, admits covered provider attempts, and creates verifiable evidence; it is effective only on fully mediated paths with authoritative shared state. | View Project → |
| ironcurtain | ★ 610 | AI Agents | TypeScript | agent-security · sandboxing · policy-engine · mcp · prompt-injection · constitution · v8-isolate · docker · multi-agent · workflow-orchestration · human-in-the-loop · zero-trust · tool-calling · escalation · typescript · cli | IronCurtain is a TypeScript runtime that keeps autonomous AI agents untrusted: it compiles a plain-English constitution into a deterministic policy, then enforces allow/deny/escalate on every MCP tool call inside a V8 or Docker sandbox. Also runs multi-agent workflows with human gates. | View Project → |
| openrig | ★ 609 | AI Agents | TypeScript | multi-agent · agent-orchestration · coding-agents · agent-harness · CLI · tmux · YAML | OpenRig is a TypeScript CLI and harness for running Claude Code and Codex together as persistent, coordinated agent teams. Define team topologies in YAML, manage sessions through tmux, and enable agents to communicate, hand off tasks, and review work. | View Project → |
| wayland | ★ 606 | AI Agents | TypeScript | desktop-agent · local-first · multi-agent · agent-orchestration · cli-orchestration · mcp · acp · skills-library · persistent-memory · electron · workflows · scheduling · voice · self-hosted · sandboxing · byo-api-key | Wayland is a local-first desktop AI agent (Electron + Rust engine) that acts as a command center for Claude Code, Codex, Gemini and other AI CLIs. It adds persistent memory, skills/workflow libraries, self-assembling agent teams, MCP/ACP integration, scheduling and sandboxed tool execution on your own keys. AGPL-3.0. | View Project → |
| Athena-Public | ★ 589 | AI Agents | Python | AI agents · local-first · persistent memory · agent governance · personal knowledge management · GraphRAG · RAG · semantic search · model-agnostic · decision support · MCP · knowledge graph | A Python agentic personal knowledge system that stores portable context in Markdown and uses RAG, structured reasoning, and governed agents to support decisions across LLM providers. Includes an MCP server and local-first workflows. | View Project → |
| agent-safe-pipeline | ★ 589 | AI Agents | TypeScript | ai-agents · agent-permissions · agent-safety · authorization · human-in-the-loop · policy-as-code · execution-authority · audit-trail · mcp · zero-trust · intent-binding · single-use-grant · verifiable-evidence · reference-architecture · typescript | TypeScript reference architecture enforcing an authorization boundary for AI agents: agents propose immutable intents, an independent policy engine returns ALLOW/ESCALATE/BLOCK, verified human approval issues a single-use execution grant, and a SafeExecutor runs only granted actions with verifiable audit dossiers. | View Project → |
| dsh-desktop | ★ 587 | AI Agents | JavaScript | ai-agent · agentic-workflows · desktop-app · electron · mcp · deepseek · ai-workbench · plugin-marketplace · local-first · multi-agent · session-management · office-automation · git-worktree · typescript · react · web-research | DSH Desktop is an Electron desktop distribution of the DeepSeek Harness agent runtime: it bundles official DSH web UI, sessions, MCP tools and a plugin marketplace with file, Git, terminal, worktree, task-board and Office artifact bundles. | View Project → |
| hunter-community | ★ 564 | AI Agents | Python | ai-agents · multi-agent · fintech · investment-research · stock-analysis · local-first · self-hosted · byok · mcp · skills · docker-compose · quantitative-trading · llm-agent · anthropic-agent-skills | 自托管的 AI 多智能体投研终端,覆盖 A股/港股/美股行情、深度分析、龙虎榜、风险扫描与走势预测。用 SKILL(Anthropic Agent Skills 格式)承载分析方法论,支持自带大模型 key 和自接 MCP 数据源,对话与持仓全部存本地。 | View Project → |
| agent_learning | ★ 555 | AI Agents | HTML | ai-agents · tutorial · book · rag · mcp · langchain · langgraph · multi-agent · memory-system · agentic-rl · context-engineering · dspy · grpo · llm · education · bilingual | Bilingual (EN/ZH) 23-chapter AI Agent textbook covering tools, memory, planning, RAG, context engineering, agentic RL, multi-agent systems, evaluation, security, and deployment, plus a small runnable reference agent with MCP server and offline tests. | View Project → |
| livecontext-ce | ★ 548 | AI Agents | Java | AI agents · AI automation · workflow automation · self-hosted · low-code · no-code · scoped agents · agent tools · audit trails · browser agent · web search · MCP | Self-hosted AI automation platform that turns chat instructions into readable workflows, scoped agents, and small team apps. Includes agent auditing, built-in data tables, and optional browser and web-search services. | View Project → |
| easyeda-agent | ★ 547 | AI Agents | Go | EasyEDA Pro · EDA automation · PCB design · schematic design · agent skill · MCP server · CLI · hardware design · circuit design · design rule checks | Automates EasyEDA Pro through typed schematic and PCB actions exposed as a CLI, agent skill, and stdio MCP server. Agents can create and validate circuit designs, manage component libraries, and inspect design-rule checks. | View Project → |
| caura | ★ 539 | AI Agents | Python | agent-memory · multi-agent · mcp · knowledge-graph · multi-tenant · pgvector · semantic-search · vector-search · llm-memory · trust-tiers · audit-trail · governance · rag · self-hosted · python · fastapi | Governed shared memory layer for AI agent fleets: agents write plain text, Caura stores it with semantic search, entity/knowledge graphs, per-tenant isolation, trust tiers and audit trails. MCP-native plus REST and Python/TypeScript clients; self-hostable via Docker with Postgres/pgvector. | View Project → |
| Agenvoy | ★ 535 | AI Agents | Go | ai-agents · mcp · mcp-server · mcp-client · self-hosted · go · single-binary · sandbox · tool-generation · multi-agent · subagents · memory · scheduler · tui · voice · llm-router | Self-hosted AI agent harness shipped as one Go binary: it plans multi-step tasks, executes sandboxed tools, and can write, test and repair new tools itself. Exposes those tools over MCP so Claude Code, Codex and other clients can share one tool library. | View Project → |
| sentient-os | ★ 524 | AI Agents | Swift | on-device-llm · local-first · privacy · macos · computer-use · mcp · personal-ai · knowledge-base · proactive-agent · ai-agents · apple-silicon · second-brain · zero-access-encryption · gemma · swift · litert | macOS personal AI agent: an on-device LLM (Gemma via a LiteRT-LM fork) reads your files, messages and email nightly into a local markdown knowledge base, then drafts proactive morning actions. Adds a computer-use Sidekick and an opt-in zero-access-encrypted MCP server for ChatGPT/Claude. | View Project → |
| freecad-ai | ★ 523 | AI Agents | Python | freecad · cad · 3d-modeling · llm · agent · tool-calling · mcp · skills · python · workbench · ollama · local-llm · multi-provider · vision · code-generation · parametric-modeling | FreeCAD AI is an agentic assistant workbench for FreeCAD that turns natural-language prompts into 3D models by generating and executing Python, with Plan/Act modes, 50 tool-calling CAD operations, skills, MCP support and 21 LLM providers. | View Project → |
| hermes-plugins | ★ 519 | AI Agents | Python | plugins · autonomous-agent · hermes-agent · model-routing · cost-control · observability · agent-memory · self-improvement · prompt-injection · hallucination-detection · sandboxed-execution · telemetry · telegram-bot · mqtt · delegation · scheduler | A collection of 23 Python plugins that extend the hermes-agent autonomous agent with goal management, multi-model routing and debate, cost guardrails, telemetry, memory consolidation, and safety filters. Install by copying plugin folders into ~/.hermes/plugins/. | View Project → |
| stagehand-python | ★ 516 | AI Agents | Python | browser automation · computer use · AI agents · web automation · natural language · Python SDK · remote browser · self-healing · workflow automation | Python SDK for Stagehand’s AI browser automation API. It lets applications observe, act, extract data, and run multi-step browser tasks using natural-language instructions alongside Playwright and code. | View Project → |
| JIT | ★ 507 | AI Agents | Python | meta-agent · agent harnesses · harness generation · harness evolution · task-specific agents · agent evaluation · multi-benchmark | JIT-Agent is a meta-agent that generates and revises executable, task-specific harnesses for agentic LLMs. It includes harness designs, benchmark adapters, and candidate selection for evaluating workflows across research, office, planning, and other tasks. | View Project → |
| agent-craft | ★ 491 | AI Agents | Python | tutorial · langchain · langgraph · mcp · rag · multi-agent · function-calling · streamlit · agents-sdk · swarm · embeddings · chroma · faiss · human-in-the-loop · chinese · beginner-guide | Chinese-language, hands-on tutorial repo that teaches building AI agents end to end: LLM calls, function calling, LangChain, RAG with FAISS/Chroma, LangGraph workflows, MCP servers and clients, Agents SDK/Swarm multi-agent, and a Streamlit UI. 13 runnable modules, each with its own code and in-depth blog write-up. | View Project → |
| QuickVoice | ★ 489 | AI Agents | TypeScript | voice agents · phone agents · self-hosted · telephony · LiveKit · Twilio · Telnyx · inbound calls · outbound calls · RAG · MCP tools · call automation | Self-hostable platform for building and operating AI phone agents, with a LiveKit voice runtime, Twilio and Telnyx integrations, a customer console, call campaigns, knowledge retrieval, and MCP tool connections. Requires provider credentials for live calls. | View Project → |
| Dex | ★ 485 | AI Agents | Python | AI assistant · chief of staff · personal knowledge management · Claude Code · Cursor · MCP · productivity · meeting notes · task management · local files | A personal AI chief-of-staff system that works through Claude Code or Cursor with a user-owned notes vault. It provides workflows for planning, meeting preparation, task follow-up, and personal knowledge management, with optional connected services. | View Project → |
| awesome-agent-native-services | ★ 476 | AI Agents | Python | AI agents · agent infrastructure · service directory · agent skills · MCP · browser automation · agent memory · sandboxing · payments · LLM observability | A curated catalog of agent-native services and infrastructure, with selection criteria and onboarding guides. Includes installable skills that help agents find, evaluate, and connect to services across categories such as email, browsers, memory, sandboxes, and MCP tools. | View Project → |
| ratel | ★ 461 | AI Agents | Rust | context engineering · tool selection · skill retrieval · agent memory · BM25 · semantic search · progressive disclosure · token optimization · tool calling | Rust retrieval engine with TypeScript and Python SDKs that helps AI agents discover relevant tools, skills, and facts on demand. Uses BM25 by default, with optional semantic and hybrid search, to reduce context and avoid loading every capability up front. | View Project → |
| CoWork-OS | ★ 459 | AI Agents | TypeScript | agentic-os · local-first · multi-provider · mcp · electron · desktop-app · automation · memory · guardrails · byok · coding-agent · chat-channels · artifacts · cli · self-hosted · personal-ai | CoWork OS is an MIT-licensed, local-first agentic workspace (Electron GUI plus cowork CLI) that runs coding, knowledge-work, browser and document tasks across bring-your-own model providers, with MCP connectors, skills, channels, memory and governed approval profiles. | View Project → |
| marketing-team-eve-template | ★ 444 | AI Agents | TypeScript | multi-agent · marketing · delegation · human-in-the-loop · Slack · Notion · Typefully · Resend · content creation · SEO | A TypeScript template for a lead agent that delegates marketing work to five specialists for content, social, SEO, product messaging, and email. Agents work with Notion, Typefully, and Resend, with human approval required for consequential actions. | View Project → |
| smithers | ★ 421 | AI Agents | TypeScript | agent-orchestration · multi-agent · workflow-engine · durable-execution · human-in-the-loop · background-agents · typescript · coding-agents · developer-tools · claude-code · codex · llm · react · cli · code-review · repo-automation | Smithers is a TypeScript agentic workflow framework: you declare flows next to your code and it runs coding agents through plan, execute and review steps with durable execution and human-in-the-loop gates. Ships a CLI, libraries and a web UI for automating repo changes. | View Project → |
| ai-platform-engineering | ★ 421 | AI Agents | TypeScript | multi-agent-systems · agent-orchestration · agentic-workflows · ai-governance · mcp · agent-skills · platform-engineering · kubernetes · cloud-native · workflow-automation · observability · devops · enterprise-ai · ai-platform · deepagents · langgraph | CAIPE is an open-source platform for building, deploying, and governing AI agents and agentic workflows. It combines multi-agent orchestration (LangGraph/DeepAgents), MCP tool integrations, agent skills, identity/authorization, and observability, targeting platform engineering and enterprise DevOps use cases. | View Project → |
| rea | ★ 417 | AI Agents | TypeScript | reverse-engineering · binary-analysis · static-analysis · decompiler · disassembler · MCP-server · agent-skills · CLI · JavaScript-analysis | REA is a CLI and MCP server that lets coding agents inspect applications, JavaScript behavior, and native binaries using tools such as Hopper or Ghidra. It returns structured evidence and investigation results to help explain features and guide reimplementation. | View Project → |
| ai-security-tool | ★ 413 | AI Agents | HTML | cybersecurity · security agents · vulnerability scanning · CVE · SBOM · penetration testing · CTF · DevSecOps · terminal · MCP | Cross-platform security terminal and vulnerability scanner with autonomous agents for audit, reconnaissance, CTF, and CVE workflows. Includes SBOM analysis, multi-protocol remote access, and an AI assistant supporting DeepSeek, OpenAI, and custom APIs. | View Project → |
| typesafe-mario | ★ 394 | AI Agents | Python | game-playing agent · computer-use · structured state · emulator control · Super Mario Bros. · direct control · TypeSafe · Jev | A Python agent uses TypeSafe's Jev model to choose controller actions for Super Mario Bros. It sends structured emulator state—including player motion, hazards, and terrain—to the model, then logs decisions and game outcomes. | View Project → |
| jarvis-assistant-vocal | ★ 392 | AI Agents | Python | voice-assistant · wake-word · whisper · ollama · claude · mcp-server · home-automation · philips-hue · obs · twilio · playwright · tts · local-llm · raspberry-pi · browser-automation · long-term-memory | Local French voice assistant on Windows: wake word, Whisper STT, LLM tool-calling via OpenAI/Claude or offline Ollama, TTS. Controls Hue, OBS, PC, calendar, browser, Twilio calls, and exposes tools as an MCP server for clients like Claude Desktop. | View Project → |
| prismor | ★ 384 | AI Agents | Python | agent-security · guardrails · prompt-injection · policy-engine · mcp-gateway · human-in-the-loop · secret-prevention · supply-chain-security · tool-call-interception · self-hosted · observability · audit-trail · devsecops · agent-governance · iam · cli | Self-hosted runtime control plane that intercepts AI agent tool calls and enforces user-written policy: observe, require human approval, or block secret leaks, prompt injection, and supply-chain risk. Agent-agnostic with hooks for Claude Code, Codex, LangChain and others plus an MCP gateway and LLM proxy. | View Project → |
| pizza-bot | ★ 383 | AI Agents | TypeScript | ai-agents · local-first · human-in-the-loop · langgraph · deepagents · mcp · agent-skills · electron · desktop-app · cli · long-running-tasks · approval-queue · checkpointing · subagents · plugins · multi-provider | Pizza Bot is a local-first desktop inbox for long-running DeepAgents/LangGraph AI agents: tasks run in the background, finished work collects in Unread and approvals in Action. Supports MCP servers, agent skills/subagents, checkpointed resumable runs, and many model providers. | View Project → |
| EDDI | ★ 380 | AI Agents | Java | multi-agent · orchestration · config-driven · conversational-ai · quarkus · java · langchain4j · mcp · a2a · rag · persistent-memory · enterprise · compliance · self-hosted · docker · observability | Java/Quarkus middleware that turns versioned JSON configs into multi-agent conversational AI systems. Adds routing, persistent memory, RAG, MCP/A2A protocols, audit ledger and GDPR/HIPAA/EU AI Act controls. Self-hostable via Docker with 12+ LLM providers. | View Project → |
| agent-second-brain | ★ 372 | AI Agents | Python | personal assistant · second brain · voice assistant · knowledge graph · memory system · Telegram bot · Obsidian · Claude Code · reminders · self-hosted | A self-hosted Telegram personal assistant that uses a persistent Claude Code session to turn voice notes, messages, and attachments into linked Obsidian knowledge. It supports knowledge-graph maintenance, natural-language reminders, and scheduled processing. | View Project → |
| rrsi | ★ 371 | AI Agents | Python | agent harnesses · self-improvement · agent optimization · benchmarking · evaluation · harness evolution · coding agents · multi-domain agents | A Python system that iteratively proposes, screens, evaluates, and selects changes to AI agent harnesses, including prompts, tools, memory, and control flow. It supports coding, document-work, and engineering-design benchmarks, with auditable Git-based candidate histories. | View Project → |
| kition | ★ 371 | AI Agents | TypeScript | AI agent · desktop workspace · browser research · tool use · document editing · visual workflows · whiteboard · structured data · local-first · MCP | Kition is an Electron workspace with a tool-using AI agent that researches in browser tabs and works with Markdown documents, structured tables, and whiteboards. It supports reviewable edits, AI-assisted table fields, and visual automation workflows. | View Project → |
| autolith | ★ 369 | AI Agents | Common Lisp | terminal-agent · self-modifying · common-lisp · sbcl · lisp-repl · mcp · coding-agent · sub-agents · recursive-inference · nix · image-recovery · checkpoints · multi-provider · skills · cli | Autolith is a self-modifying Common Lisp terminal AI agent living in a live SBCL image: the prompt is a REPL, it exposes self/lisp tools, sub-agents, MCP, and recursive inference (RLM) over oversized inputs, with checkpoints and overflow to a recovery image. | View Project → |
| agentic-engineering-handbook | ★ 367 | AI Agents | Python | agent engineering · learning roadmap · tutorials · MCP · agent harnesses · evaluations · tool use · memory · skills · coding agents · multi-agent systems | A curated learning roadmap with 181 resources, tutorials, and exercises covering AI agent loops, MCP, context and memory, skills, harnesses, evaluations, and production systems. | View Project → |
| clearai-dsh | ★ 366 | AI Agents | JavaScript | agent-preset · dsh-plugin · deepseek-harness · epistemic-loop · ontology · knowledge-graph · scientific-discovery · evidence-tracking · hypothesis-tracking · verification · local-first · research-assistant · memory · skills · nodejs · agent-framework | ClearAI is a DSH (DeepSeek Harness) agent preset that runs a seven-stage epistemic loop — frame, hypothesize, plan, observe, verify, evaluate, record — to grow a verifiable domain ontology. Each knowledge edge needs evidence and independent evaluation, with contradictions surfaced for human decision. | View Project → |
| awesome-OpenClaw-Money-Maker | ★ 358 | AI Agents | JavaScript | awesome-list · AI agents · OpenClaw · agent monetization · agent skills · trading agents · workflow automation · MCP · crypto | A curated directory of open-source tools, agents, skills, and strategies for earning income with AI agents such as OpenClaw. Covers trading, marketing, content creation, workflow automation, and MCP integrations. | View Project → |
| exxperts | ★ 358 | AI Agents | TypeScript | local-first · agents · memory · approval-gated · mcp · multi-provider · privacy · desktop-app · typescript · web-search · persistent-rooms · skills · no-telemetry · self-hosted · background-tasks · artifacts | Local-first AI agent app (desktop, web, CLI) with governed, approval-gated memory stored as plain files on your machine. Any model provider (Claude, ChatGPT, Gemini, Mistral, OpenAI-compatible), built-in MCP tools and web search, no cloud or telemetry. | View Project → |
| awesome-deepseek-harness | ★ 353 | AI Agents | TypeScript | awesome-list · DeepSeek Harness · AI agents · coding agents · MCP · agent skills · plugins · orchestrators | A curated directory of plugins, skills, MCP servers, patches, orchestrators, and UIs for DeepSeek Harness, covering coding, visualization, presentations, and automated research. | View Project → |
| agentica | ★ 352 | AI Agents | Python | multi-agent · agent-framework · cli · python-sdk · persistent-memory · self-evolving-skills · mcp · rag · workflow · swarm · async · context-compression · guardrails · gateway · im-integration · desktop-app | Agentica is an Apache-2.0 Python SDK and CLI for building multi-agent systems: each terminal session is a collaborating agent, with async tools, persistent memory, self-evolving SKILL.md, MCP/RAG support, and a local Web/Desktop gateway plus WeChat/Feishu callbacks. | View Project → |
| mulmoclaude | ★ 350 | AI Agents | TypeScript | local-first · personal-assistant · ai-assistant · claude-code · mcp · chat-ui · plugins · knowledge-management · self-hosted · no-code · personal-wiki · privacy · schema-driven · automations · bridges · typescript | Local-first, MIT personal AI assistant platform built on Claude Code: memories, schema-driven collections, feeds and mini-apps stay as plain files in your workspace. Chat summons the right GUI (wiki, charts, forms, spreadsheets, 3D), capabilities are plugins, and 20+ messaging bridges relay to the same self-hosted agen | View Project → |
| mobile-jev | ★ 341 | AI Agents | JavaScript | android-agent · computer-use · mobile-automation · device-control · ui-automation · agent-loop · cli · react-studio · execution-traces · latency-profiling · observability · typesafe-jev · mobilerun · nextjs · typescript · mit | Mobile Jev is a standalone Android computer-use agent: TypeSafe's Jev picks one typed operation per observation (OPEN_APP, TAP, TYPE_TEXT, DONE, BLOCKED) and Mobilerun executes it on a real phone. Ships a React studio, CLI with traces and latency profiling, and no ADB requirement. | View Project → |
| stash | ★ 338 | AI Agents | Python | agent memory · continual learning · trace collection · knowledge base · skills · semantic search · multi-agent · MCP · coding agents · CLI | Stash captures agent traces and human corrections, curates durable knowledge, and makes it retrievable through a CLI, API, MCP, and skills. It is infrastructure for carrying lessons across agent runs without changing model weights. | View Project → |
| agents-universe | ★ 336 | AI Agents | Python | AI agents · multi-agent · agent orchestration · shared project context · agent skills · workflows · enterprise · auditing · self-hosted · scheduled tasks | Self-hosted enterprise AI agent platform with built-in role-based agents, shared project context, skills, and workflows. Agents can plan and execute tasks, create pull requests, run tests, and record auditable work; users configure supported LLM providers. | View Project → |
| comanda | ★ 325 | AI Agents | Go | cli · yaml · workflow-orchestration · multi-agent · coding-agents · agentic-loops · quality-gates · mcp · semantic-memory · knowledge-graph · codebase-index · git-worktrees · terminal · automation · stateful-checkpoints · go | Go CLI that turns English goals into inspectable YAML agent workflows, coordinating Claude Code, Codex, Gemini CLI, Kimi Code and API/local models. Adds stateful agentic loops, quality gates, codebase indexes/knowledge graphs, semantic memory, and an MCP server. | View Project → |
| oya-browser | ★ 324 | AI Agents | TypeScript | AI agents · browser automation · computer use · MCP · web automation · browser control · Playwright · Puppeteer · CAPTCHA handling · MFA · session persistence · workflow replay | A browser control plane and TypeScript SDK for AI agents, with MCP support and connections to managed or self-hosted Chrome. Agents can automate authenticated workflows, handle human handoffs for CAPTCHA or MFA, and save tasks as replayable playbooks. | View Project → |
| dexter-jp | ★ 308 | AI Agents | TypeScript | ai-agent · finance · japanese-stocks · edinet · jquants · langchain · research-agent · autonomous-agent · cli · typescript · rag · tool-use · multi-source · agent-skills · disclosure-analysis · slack-discord-bot | Dexter JP is a TypeScript autonomous research agent for Japanese listed companies, built on LangChain and the Claude Agent SDK. It plans, calls EDINET DB and J-Quants tools in parallel, checks hypotheses against filing paragraphs with citations, and outputs structured financial reports via CLI or chat gateways. | View Project → |
| 0 | ★ 307 | AI Agents | TypeScript | AI security · autonomous agent · penetration testing · vulnerability research · code review · exploit verification · security remediation · MCP · multi-model · DevSecOps | 0 is a local, multi-model AI security agent for investigating authorized codebases and applications. It can review source, run security tools, examine findings, and propose fixes through an extensible toolchain. | View Project → |
| DashClaw | ★ 307 | AI Agents | TypeScript | agent-governance · approvals · human-in-the-loop · policy-engine · mcp · mcp-server · hooks · claude-code · codex · audit-ledger · agent-runtime · prompt-injection · unattended-agents · self-hosted · security · ai-ops | Self-hosted governance layer that intercepts AI agent tool calls before they run, routing destructive or expensive actions to remote human approvals and recording signed decision evidence. Integrates via Claude Code/Codex/Hermes hooks, an OpenClaw gateway, an MCP server, and Node/Python SDKs. | View Project → |
| agentao | ★ 306 | AI Agents | Python | python · agent-runtime · local-first · mcp · acp · permissions · memory · audit-replay · cli · embeddable · openai-compatible · sub-agents · skills · sandbox · plan-mode · session-replay | Agentao is a local-first, embeddable Python agent runtime: use it as a library, CLI, or ACP server. It ships permission modes, tool confirmation, MCP and skills, SQLite memory, sub-agents, and JSONL audit replay with no cloud dependency. | View Project → |
| deepdoc | ★ 305 | AI Agents | Python | deep research · local knowledge base · document analysis · multi-agent · evidence verification · Qdrant · report generation · RAG | A local deep-research agent that indexes PDF, DOCX, PPTX, image, text, and Markdown files in Qdrant, researches report sections in parallel, and writes a Markdown report. Jev evaluates evidence relevance, coverage, contradictions, and injection-like content. | View Project → |
| home-generative-agent | ★ 305 | AI Agents | Python | home-assistant · smart-home · ai-agent · langgraph · anomaly-detection · computer-vision · face-recognition · pgvector · semantic-memory · local-llm · ollama · voice-assistant · hacs · rag · automation · proactive-alerts | Home Assistant integration with a LangGraph agent that controls entities, writes real HA automations from plain English, analyzes cameras with face recognition, and runs local pgvector memory plus a deterministic Sentinel anomaly/security engine. Works with OpenAI, Gemini, Anthropic, or fully local Ollama. | View Project → |
| Cotal | ★ 303 | AI Agents | TypeScript | multi-agent · agent-coordination · pub-sub · agent-mesh · a2a · mcp · nats · jetstream · agent-skills · presence · durable-delivery · channels · orchestration · self-hosted · cli · distributed-systems | Cotal is an open pub/sub standard and TypeScript reference implementation for coordinating many AI agents in one shared space over NATS/JetStream. It provides presence, channels, durable unicast and role anycast addressing, a web dashboard, plus connectors for Claude Code, Codex, OpenCode and Hermes. | View Project → |
| moai-cowork | ★ 302 | AI Agents | Python | AI agents · AI coworkers · multi-agent · Korean · Claude Cowork · ChatGPT Work · Codex · agent marketplace · no-code · MCP · skills · business workflows | A marketplace of Korean-language specialist AI coworker plugins for Claude Cowork and ChatGPT Work. Its PM agent sets up projects and routes natural-language requests through role-specific agents and skill chains for business, creative, and professional workflows. | View Project → |
| openmuse | ★ 299 | AI Agents | TypeScript | personal-agent · computer-use · browser-agent · playwright · terminal · task-worker · ag-ui · copilotkit · react-native · expo · human-in-the-loop · gmail · google-calendar · pdf · self-hosted · typescript | OpenMuse is a self-hosted personal AI agent built on CopilotKit and AG-UI that combines a persistent Chromium browser, an optional Docker Linux terminal and workspace, files/PDFs, Gmail and Calendar access, and a durable task worker with human review of actions. It ships as a MIT-licensed TypeScript monorepo with React | View Project → |
| chmonitor | ★ 297 | AI Agents | TypeScript | ClickHouse · AI agent · MCP · database monitoring · observability · query optimization · self-hosted | Self-hosted ClickHouse monitoring dashboard with an AI agent and MCP interface for investigating cluster health and getting index, partition, and materialized-view recommendations. Recommendations are advisory; the tool does not apply DDL. | View Project → |
| plur | ★ 296 | AI Agents | TypeScript | agent memory · shared memory · persistent memory · local-first · MCP · hybrid search · embeddings · memory retrieval · context management · open format | PLUR provides local-first, persistent memory shared across AI agents and coding tools. It stores inspectable text-based memories and supports MCP integrations, scoped retrieval, and hybrid search with local embeddings. | View Project → |
| Context-Engineering-for-Multi-Agent-Systems | ★ 283 | AI Agents | Jupyter Notebook | multi-agent-systems · context-engineering · mcp · rag · agentic-ai · semantic-orchestration · observability · glass-box · langchain · nemotron · pinecone · gradio · prompt-injection-defense · token-analytics · computer-use · governance | Book companion repo (Jupyter notebooks) that builds a domain-agnostic, glass-box Context Engine for multi-agent systems: MCP orchestration, dual RAG with citations, injection defenses, and token analytics, runnable on OpenAI, LangChain, Nemotron, or DeepSeek-R1. | View Project → |
| Containarium | ★ 282 | AI Agents | Go | agent-runtime · agent-sandbox · isolation · mcp · ebpf · kubernetes · lxc · ssh · self-hosted · multi-tenant · gpu-passthrough · sandboxing · coding-agents · go · code-execution · egress-policy | Self-hosted agent runtime that gives each AI agent a persistent, SSH-reachable isolated box on LXC or Kubernetes, with eBPF egress policy, GPU passthrough, and MCP-native CLI plus in-box shell/file tools for Cursor, Claude Code, and other MCP clients. | View Project → |