GitHub agents

All the code files shared in the Vizuara AI Agents 10 day bootcamp

optimize your documentation through fleets of agents

ShakaCode Team Shared Claude Code Commands, Agents, Skills, Tips

This is the repo for the paper A Self-Evolving Framework for Efficient Terminal Agents via Observational Context Compression

An Agentic Data Preparation Framework for AGI-driven Scientific Discovery

智能终端命令助手 — 用自然语言生成 Shell 命令，支持 OpenAI/Claude/Gemini/Ollama 等多后端

A curated list of tools, frameworks, patterns, and research for AI-driven and agentic approaches to modernizing legacy systems

PiloTY: AI pilot for PTY operations via MCP - enables AI agents to control interactive terminals like a human

Production-grade multi-agent orchestration engine. Compose agentic workflows from a pluggable catalog of Agents, LLMs, tools, and retrievers. Execute locally with LangGraph or distributed with Temporal. Built-in RAG pipeline for enterprise knowledge retrieval. A2A and MCP protocol support. Visual drag-and-drop blueprint builder.

Personal AI agent harness powered by Claude Code. Always-on access through Discord, Telegram, or a built-in web UI.

Multi-agent workflow for accurate and comprehensive literature reviews in philosophy

Flutter Debug Bridge - CLI for AI agents to interact with running Flutter apps on device

An AI agent that controls your Flipper Zero that uses PyFlipper, LightLLM, RAG, and can connect to your local Ollama instance!

Give your AI agent eyes and hands on any desktop — cross-platform accessibility API with MCP server

AI Agent Skills for Google Play Console CLI (gplay)

BackdoorAgent is a stage-aware framework and benchmark that instruments LLM-agent workflows (planning, memory, tools) to systematically inject, track, and evaluate backdoor triggers across multi-step trajectories using unified metrics like ASR and clean accuracy.

A miniature isometric 3D office where 9 AI agents work, walk, talk, and hold meetings. Built with React 19 + Three.js.

Claude Code skill for handling long-context tasks through recursive decomposition

A platform evolution of OpenAI Symphony (openai/symphony) for running Codex, Claude Code, CodeBuddy Code, and OpenCode from real project tasks.

LangGraph Typescript Agents Notebooks: email, human in the loop, memory

A Kanban-style TUI for parallel AI coding. Manage Git worktrees as tasks, run multiple agents in isolated tmux sessions, get notified via hooks. Built for developers who vibe with Claude Code, Cursor, and all other agents..

Rails-native LLM cost ledger: track spend by provider, model, and feature with self-hosted storage and budget guardrails.

Workshop series teaching the AI Engineering Lifecycle using LangChain, LangGraph, and LangSmith

The self-evolving agentic framework for bioinformatics

🕵️‍♂️ Privacy-focused AI job scraper, local storage, and interactive dashboard. Auto-scrapes AI/ML roles from top companies using ScrapeGraph-AI + LLM and LangGraph Agents, filters for relevance, and provides a Streamlit UI for tracking applications. Built for developers seeking AI careers.

Open-source AI agent CLI built in Zig. A pure Zig rewrite of OpenClaw — one ~5MB binary, zero Node.js. Agentic loop, SSE streaming, tool dispatch, SQLite history, and multi-backend support (Claude, Ollama, OpenAI-compatible). Cross-compiles Mac, Linux, and Windows in one command.

34 FinOps specialist AI agents + 6 named-pattern playbooks. Drop into Claude Code, Copilot, Cursor, Windsurf, Aider, Gemini CLI. MIT.

SochDB is a high-performance embedded, ACID-compliant vector database purpose-built for AI agents and memory

🛡️ RakshaQuant - AI-powered paper trading for NSE using LangGraph multi-agent orchestration. Features 4 LLM agents, learning feedback loop, LangSmith observability & risk guardrails. Built with Groq, DhanHQ sandbox & PostgreSQL. Paper-first design.

CVE hunting harness for Claude Code - 20 skills, 5-agent team, systematic vulnerability research with false positive elimination

[ACL'26 Oral] AgentOCR is a token-efficient framework that compresses multi-turn agent history by rendering it into images and adopting RL-driven self-compression

Management consulting skills for Claude Code, Cowork, Codex, and other agents. Strategy, operations, and transformation workflows for the full engagement lifecycle.

AI made building cheap. It didn't make deciding cheap. Mycelium is a Claude Code harness that makes your agent run discovery and weigh evidence before it writes code. It earns the right to start. Built for software, courses, AI tools, and services.

A CLI that extracts the useful structure from source files: signatures, imports, types, variables, comments, and Markdown sections.

Example code/configs for all frameworks (CrewAI, kagent, ADK, etc.), MCP, fine-tuning, and AI gateways

Tula is an open-source collection of OpenClaw skills, configurations, and patterns designed to transform a general-purpose AI agent into a personal health intelligence assistant.

Use LLM as a logic gate/processor for local coding agent, context management customisation so the llm calls remains fast and small

《Skill 蓝皮书 2026》· 基于 61,776 条 Skill 数据的原生研究 · AgentSkillsHub 出品 · 54% 0 star · 基尼系数 0.983

Obsidian AI is an open-source platform that gives you a full visual interface for building, managing, and running AI agents — no SDKs, no boilerplate, no glue code required.

The whole local AI stack in one executable: it runs and manages local AI models across every GPU, and it's a search engine you can talk to, with cited answers from your files, code, and the web. MCP server for coding agents, web crawler, TUI, CLI, REST API, Python library. No Ollama or LM Studio needed, works with both.

A local proxy server that relays Anthropic Messages API-compatible requests to the Kiro (Amazon Q) backend using Kiro CLI credentials.

SEO blog writer Claude skill — turn any URL, notes, or topic into a human-sounding blog post that ranks on Google. Fills every SEO field: title, meta description, FAQ schema, table of contents. Beats AI detectors.

A neuro-symbolic runtime architecture (DIR) for Responsibility-Oriented Agents (ROA). By decoupling probabilistic LLM reasoning (User Space) from deterministic execution (Kernel Space), it ensures safe, invariant-driven, and auditable AI autonomy.

CodeAuditSkill 是一个运行在 Claude Code 中的 Web 项目安全代码审计/漏洞挖掘 Skill， 已经获得了40余个CNVD漏洞编号。 给定一个 Web 项目目录，自动识别语言与框架，按语言专属漏洞清单进行系统化源码审计，对每条高中危漏洞并行启动子 Agent 进行可利用性验证，最终输出可交付的审计报告。

Local semantic copilot and governed agent engineering runtime for AI-assisted software delivery.

A collection of AI agent skills for QEMU hardware modeling

Gaster Code: public local AI coding assistant with CLI, desktop app, MCP, Skills, Agents, and signed installers

Hunch it is your personal AI trading desk for tokenized stocks and crypto on Solana. As you review, skip, and execute trades, you unlock new AI trading agents, level up their capabilities, and receive increasingly personalized trade proposals with one-tap execution and built-in TP/SL.

CarryCode is a terminal-native AI coding agent that helps you write, refactor, debug, and understand code — all through natural conversation. It connects to 17+ LLM providers, supports the MCP protocol for extensibility, and delivers a beautiful terminal UI with themes, syntax highlighting, and code diff previews.

Public binary downloads for Agent Cat

Train an agent to generate high quality summaries

Anime-themed personal assistant and goal-oriented intelligent agent

Self-hosted DTLN noise suppression plugin for LiveKit Agents — no cloud API, no per-minute fees

Lint rules for your AI agent's discipline, not its code

MCP server for Git with local Ollama — zero tokens for git operations

Adversarial, deterministic benchmarking for AI agent skills — measures whether a skill makes the agent measurably better. Includes biomedical eval cases.

Agent-managed GitHub merge queue for Codex, Claude Code, Cursor, and MCP clients

Protokół dla polskiej inteligencji.

AI runtime for creating, running, and supervising background AI workers.

A local-first, privacy-preserving alternative to Google NotebookLM — Tauri 2, React 19, Rust, Ollama

Hands-on lab: Get Started with Models in Microsoft Foundry — From First Inference to Deployed Agent (Microsoft Build 2026, LAB520)

A Minimal Agent for Automated Theorem Proving

Track and analyze skill usage by AI agents, and clean up redundant or overlapping skills.

EarthLink: A Self-Evolving AI Agent System for Climate Science

Agentic Software Engineering (ASE)

Official benchmark reproduction code for the SAG paper

Notion automation CLI for AI agents

An OpenCode plugin that sandboxes agent-executed commands using @anthropic-ai/sandbox-runtime

企业知识库 RAG 智能问答系统

Elixir-native LangChain, LangGraph, and DeepAgents for traceable LLM apps: OTP workflows, tools, memory, human-in-the-loop, streaming, custom clients/adapters, minimal deps, and WeaveScope tracing.

基于 LangGraph + Flask 的 AI 剪映视频剪辑助手（Agent），通过对话式交互与剪映(CapCut)集成，使用自然语言完成智能视频制作

AI evaluators CLI for your AI apps and Agents - Dynatrace AI Observability

Build a Claude-Code-shaped agent harness from scratch. 7-week course, 20 chapters, ~5,000 lines of Python, 42 tests, 3 LLM providers, no frameworks.

Make AI coding agents safe to scale autonomously: assign work, cap spend, enforce policy, verify output, roll back failures, learn from loops, and prove ROI across every repo.

Agent Skills de conformidade LGPD (Lei 13.709/2018), resoluções ANPD e ECA Digital (Lei 15.211/2025) — instaláveis em Claude Code, Codex, Gemini CLI, Cursor e OpenCode. 1 maestro (lgpd-audit) + 18 sub-skills.

LangChain RunnableParallel example: Research a company by running API calls in parallel. Demonstrates RunnablePassthrough.assign, partial failure handling, and a canonical Pydantic schema as the retrieval layer output. Production-ready pattern for parallel data collection in LangChain.

An AI Multi-Agent memory system that remembers your code style, fixes, and preferences — for Hermes-Agent.OpenClaw,Opencode,Claude Code, etc.

AI coding agent skills for Three.js and React Three Fiber best practices

A Claude Code skill that orchestrates a live agent team for strategic decisions. A team lead runs intake, classifies the question, spawns 3-4 teammates that work McKinsey-style frameworks in parallel, synthesizes a recommendation, then a red-team teammate attacks its own logic (pressure-test) before it ships.

🚀 FullScopeTest 是一款 AI 驱动的下一代全链路自动化测试平台。深度融合大模型能力，提供自然语言生成脚本(NL2Script)、智能错误自愈、测试数据裂变及全局 AI Copilot。无缝集成 API 测试、Web 自动化(Playwright)与性能测试(Locust)，让自动化测试变得前所未有的智能与高效！🤖✨

[ICLR 2026] Code for the paper "Graph-of-Agents: A Graph-based Framework for Multi-Agent LLM Collaboration"

Pi Desktop — a free, open-source Electron GUI for the Pi coding agent. Chat, edit files, run a terminal, and review diffs in one window. Builds for Linux, macOS, and Windows.

An 18 notebook course that isolates and measures each component of agentic loop engineering on real, industry standard software datasets.

PC-side companion for the Hermes-agent Android app — mesh (run tasks on your computer) + handoff (move conversations between PC and phone), NaCl-encrypted, pair-once, no cloud. AGPL-3.0.

X (Twitter) marketing skills for Claude Code and Codex: write tweets, threads, and replies in your voice, strip AI tells, and publish via Publora. Open source, MIT. Content engineering by Creative Content Crafts.

Proof-driven Claude/Codex agent loop with paired execution, persistent resume, and adversarial security/performance gates.

A governance framework that establishes memory as a first-class system component with explicit policies, tiered credential management, lifecycle controls, and full observability

A Dynamic Island for your coding agents — a native macOS notch monitor for Claude Code, Grok, Copilot and more. Monitor, approve, and jump back without leaving your flow.

Sistema de aprendizaje continuo para Claude Code: hooks que observan patrones de trabajo reales y los destilan en reglas atómicas con nivel de confianza, aplicadas en sesiones posteriores.

📝 Document writing skills for AI agents (Claude Code / Cursor / CodeBuddy) — 4-stage workflow: Brainstorm → Plan → Execute → Review. Inspired by obra/superpowers. GitHub 44★.

Diffium is a CLI tool to help you better understand git diffs, with the main intention to make co-development with AI coding agents easier.

全栈开发 AI Agent Skills，包含 Vue3 管理后台、UniApp 移动端、6 大后端框架（Spring Boot、NestJS、Gin、Django、ThinkPHP、ASP.NET Core）及 MySQL 数据库设计规范。

Hunch it is your personal AI trading desk for tokenized stocks and crypto on Solana. As you review, skip, and execute trades, you unlock new AI trading agents, level up their capabilities, and receive increasingly personalized trade proposals with one-tap execution and built-in TP/SL.

Hermes Agent skill for typography-led lyric videos using GPT Image 2, Gemini Omni Flash, and exact-text QC

Provide different cloud products MCP Server tools to help developers manage cloud resources with AI-agent

A Notion-like wiki for non-technical teams, backed by real git commits and a merge-request review workflow — plus AI agents that write the first draft for you.

基于 CesiumJS 1.142 的原生 Cesium Agent Skill（Cursor / Codex / Claude）。含约 185 个 JS 示例与材质脚本，覆盖地形、3D Tiles、特效等；仅 md/js，无 HTML。复制到 .cursor/skills、.agents/skills 或 .claude/skills 即可使用。

Roslyn-based MCP server giving AI agents deep semantic understanding of .NET/C# codebases — 67 tools for navigation, call graphs, diagnostics & code fixes, safe refactoring, code-quality auditing, test intelligence, DI graphs, and IL/external-assembly inspection.

The open-source agent platform to build, distribute, and optimize AI agents — on infrastructure you own.