Code agents

Local MCP code-intelligence for AI coding tools: surgical search/read/edit, agent-config version history, path-level access + masking guards, and a multi-project hub.

Self-hosted browser-backed web_search and web_fetch for Codex, Claude Code, OpenCode, Pi, OpenClaw, HermesAgent, LangChain, and custom Agents — MCP + REST with provider fallback and SSRF protection.

AetnaMem /ET-nuh-mem/ is agent Memory delivers fully local long-term auditable, proof-ready memory for AI Agents via a 4-tier progressive pipeline, with zero external API dependencies, it supports OpenClaw, install it via pip.

Cut your Claude API bill: local proxy that auto-routes every request to right model. Zero code changes.

養成型 AI 伴侶框架｜Raise an AI companion who remembers, reflects, and grows with you.

Desktop client for the Asale token-sharing network: sell idle Claude, ChatGPT/Codex, Gemini, Kimi and Grok subscription quota for USDT, or buy capacity the moment you hit a rate limit.

Curso de AI Engineering en Rust para Jeresoft Academy: RAG, herramientas, agentes, seguridad y evaluación con evidencia reproducible.

面向 AI Agent 的规范化开发 Skill，帮助构建高质量智能体 | A structured skill for building reliable AI agents with specifications, workflows, and acceptance.

ARCHIVED. Successor: https://github.com/jperezduerto/arsenal — practitioner methodology for offensive security, written in a form a language model can load.

claude2codex — bridge AI coding sessions between OpenAI Codex CLI and Anthropic Claude Code CLI. Convert JSONL sessions bidirectionally and resume where you left off.

Run Claude Code on an unrooted Android phone via Termux + proot-Ubuntu, with free model access through a local 9Router proxy.

Autonomous AI agent framework — LangGraph brain, swarm orchestration, Arabic-first, with human-in-the-loop safety.

AI software-delivery control plane for Claude and Codex. Build and review with dependency-aware readiness, scoped execution, fingerprinted evidence, 25 engineering lenses, and human approval.

PaperCraft — From idea to accepted paper at top CS conferences. 17 rules + 12 skills that guide CS researchers through the full research pipeline.

Agent skills for OpenCode, Claude Code and other AI agents

Keep GPT and DeepSeek conversations continuous in Codex Desktop, allowing the two models to work alternately within the same project. 在 Codex 桌面应用中实现 GPT 与 Deepseek 任务上下文双向交接，让两个模型在同一个项目下交替工作。

A virtual software company for Claude Code: 19 agents, 38 workflows, 16 quality gates. Takes an application project from idea to release.

Marblo — the live orchestrator for AI-native teams

nui is a self-hosted agent runtime. Define agents declaratively in ADL, run them on any harness or model, and reach them from a UI, CLI, API, or MCP.

Run AI agents in the Polkadot app. E2E-encrypted and private.

Apple-native cloud-hybrid RAG client for PDFs and personal files using SwiftUI, Vision OCR, OpenAI embeddings and Responses API, and Pinecone vector search.

Multi-agent LLM trading system implementing arXiv:2510.04787: agents develop grid-trading strategies offline and a low-latency bot executes them on Binance spot. Research implementation under active correction, not for live capital.

A local browser your coding agent can actually finish a job in — file uploads, cross-origin iframes, native dialogs, and downloads, all via CLI/MCP.

Skill that delegates frontend, coding, review, and image tasks to the local Grok CLI

Turn your Claude subscription into a multi-tenant agent API with rate limits, cost tracking, autonomous objectives, and an admin dashboard. Self-hostable, no API key.

Postgres test DB cloned straight from your production, for AI agents and humans. Real-time replication of rows and schema, copy-on-write clones, snapshots. No seeding, no dumps.

Claude Code Visual Studio extension

Korean stock market data for AI - KOSPI/KOSDAQ closing prices and DART disclosures as free JSON, plus an MCP server. No signup, no API key.

補助金申請の事業計画書づくりを、自分の Claude Code と公式公募要領でセルフサーブで回す無料 OSS テンプレート。申請代行ではありません

Agent 意图框架 — 把人话翻译成结构化意图+槽位+约束。输入规范化(两阶段) + 三层瀑布式意图识别(Code→Flash→Pro)，80% 请求零 LLM 调用，10ms 内返回。650 测试全过。

Field-tested local AI stack optimizations for HermesAgent, LM Studio, and Qwen 3 models

Agente de inteligencia artificial en Streamlit y LangChain con Google Gemini para consultas en lenguaje natural sobre inventarios, stock y proveedores en Excel.

Focusing on building both MCP servers and clients using the Python SDK to make sure these three core primitives—tools, resources, and prompts—and understand how they integrate with Claude AI, Gemini AI, Copilot and Grok to create powerful applications without writing extensive integration code.

Authoritative multi-agent software engineering coordinator with isolated Git worktrees, write scope locking, test verification, and serialized merge queue. (v0.1.0-beta).

VoxelSage turns CT scans into actionable 3D insights—automatically segmenting anatomy, quantifying tumors and vessels, generating structured reports, and powering agent-guided surgical analysis.

Open-source, local-first control plane for multi-model AI coding agents in VS Code and MCP: task DAGs, source intelligence, durable context and evidence-based review.

A credential is not authorisation. ACP moves the authorise/refuse decision for AI agent actions outside the model, where prompt injection cannot reach it. Specification, Dafny proofs, and a reference implementation — every claim replays locally.

The plunge for pi to boost his search ability

DSH Math Modeling Team Plugin Pack: 2 role agent presets (modeling-coding / paper) for DeepSeek Harness, multi-folder Git collaboration + vision sub-agent; includes a full run on 2023 CUMCM-C. ｜ DeepSeek Harness 数学建模团队插件包：2 套岗位Agent预设(建模编程+论文)、多文件夹协同+识图子代理，含2023国赛C题全流程跑通示例。

🦙 chat-o-llama: A lightweight, modern web interface for AI conversations with support for both Ollama and llama.cpp backends. Features persistent conversation management, real-time backend switching, intelligent context compression, and a clean responsive UI.

Your AI pair programmer for Visual Studio 2026 & VS Code - 100% local, no API key, no telemetry. Agentic loop with 28 tools: reads, edits, builds, tests, reviews your diff, even drives a live debug session. Ghost-text completions, semantic search, MCP, plans & background tasks, model bench/router. Ollama, LM Studio or OpenAI-compatible.

🥈 2nd Prize Winner (Expert Nomination) — Multi-agent clinical AI platform for InterSystems IRIS with FHIR R4, Vector Search RAG, LangChain, live FHIR agents, and a no-code Agent Builder.

A multi-agent engineering crew for Claude Code — one manager plans and verifies, specialist agents read, write, test, research, and review. Hard role separation, cost-tiered models, custom gateway ready.

Deterministic control plane for Codex Agents with typed delegation, durable lifecycle receipts, and journaled writer isolation.

Git-backed message bus for AI coding agents — Claude Code, Cursor and Codex coordinate asynchronously, with no server

HuntForge 铸猎 —— AI 攻防全链路自主解题 Agent（百度 Agent+ 挑战赛作品）：Claude Code 驾驶舱 + 三通道纪律调度 + 88 条离线 CVE 指纹库 + 33 本通用方法论手册 + 二进制/APK 逆向自动化（bin_triage·angr·jadx）+ 幂等提交与证据门禁

爽律 Skill｜律师专业工作流系统 — 面向中国律师实务的 Agent Skills 工作流：刑事 / 民商事 / 合同 / 法律顾问 / 尽调 / 法律研究 / 文书质控，内建对抗性审查、全链路溯源与人工复核边界。发起及主导：蔡诗爽律师。公开共创，文本 CC BY-SA 4.0 / 代码 Apache-2.0。ShuangLaw: Agent Skills workflow system for PRC lawyers.

Local-first desktop workspace for ACP agents—DeepSeek Harness, Claude Code, Codex, Gemini CLI, OpenCode, Hermes, and OpenClaw—with projects, tools, files, browser, and sessions.

A minimal terminal AI coding agent — LLM tool-use loop with sandboxed file/shell tools

MonoLight 是一个专注于安全执行与人机协同的通用自主智能体（General Autonomous Agent）运行时。

Open-source LangGraph + FastAPI starter for stateful AI agent APIs with PostgreSQL/pgvector, RAG, API-key auth, Docker, migrations, and tests.

safetensors llama Arabic English LLM Decoder

gguf code qwen qwen-coder codeqwen text-generation

把学校规章、培养方案和公告变成能提问、可追溯到原始文件与页码的大学 RAG 知识库。

Skills Board, the agent-native skills registry for teams. Free forever.

Persistent memory for Claude Code. Survives compaction and session boundaries: anti-patch reconcile-on-write with LLM-judged de-duplication, a forced PROGRESS.md handoff, a live PLAN.md anchor with an enforced directive ledger, FTS5 search, MCP tools and a GUI. Pure stdlib Python, zero runtime dependencies.

Claude Code records what you work on into Jira, Linear, GitHub Issues, Trello or Vikunja, automatically

Self-hosted Qwen3 retrieval server with Embedding, Reranker, MCP, and Web console.

Portable contracts and local-first runtimes for canonical data, retrieval, durable execution, and agent-facing AI.

Portable one-click Windows launcher for DeepSeek Harness (DSH). No Node.js, npm, pnpm or setup required. DeepSeek Harness 免安装 Windows 一键启动器。

可观察、可干预的中文 AI 写作工作台｜Observable AI writing workspace with orchestration, memory, evaluation, and feedback.

Stable, editable PPTX generation for AI agents — semantic IR in, native DrawingML out. DSH plugin + Claude Code plugin + CLI. | 给 AI agent 的稳定可编辑 PPTX 生成：语义 IR 进，原生 DrawingML 出。DSH 插件 / Claude Code 插件 / CLI。

A modern, responsive AI chat bot application supporting multiple AI providers including Google Gemini, Groq, OpenRouter, Hugging Face, and OpenAI and enable to store the chat history. Built with React, TypeScript, Vite including typewriter effect, animated icons, application insights & performance dashboard for the best user experience

Ready-to-use, customizable WhatsApp AI GPT4o Chatbot in C# for your Business. Supports GPT-4o with text + audio + image input, audio responses, and improved RAG + MCP Tools 🤩

🚀 AI-powered freelance gig aggregator MCP server for Claude Desktop. Search gigs, generate proposals, negotiate rates & optimize profiles with ChatGroq AI.

从《矛盾论》《实践论》完整方法论蒸馏的中文 AI 产品经理（AI Product Manager）/ Product Decision Agent Skill：解决需求、增长、运营、数据与项目协作；支持 Codex、Claude Code、Cursor

Autonomous multi-agent crypto intelligence & trading ecosystem for Robinhood Chain L2 (EVM #4663) operated via Discord Command Center, Terminal TUI, REST API, & Telegram.

Local-first AI character companion framework: emotion × energy × affinity state machines, built-in memory pipeline (facts, vectors, day summaries), background reflection, and proactive speech after silence—architecture-first “persona bones,” not prompt-only chat. Electron/Vue + FastAPI; optional VLM & TTS. AGPL-3.0.

A股多智能体量化投研系统 | 5 Agent深度分析 + 多空辩论 + 决策复盘 | React股票站 + Streamlit + Windows桌面 + iOS | 免费AkShare/腾讯行情数据源

Open-source AI work assistant. Connects to Gmail, Slack, Calendar, Notion, Linear and GitHub, runs autonomous scheduled workflows, and remembers what matters. Self-hosted, provider-neutral.

AI film preproduction Skill for Codex: reference-video analysis, scripts, detailed storyboards, visual assets, Seedance 2.5/2.0 prompts, and verified releases.

Native, local-first coding-agent desktop app in Rust (egui) — run any model: local GGUF, Ollama/LM Studio, SSH-tunneled runtimes, hosted APIs, or Claude Code / Cursor / Codex over ACP. No Electron, no cloud lock-in.

Final Year Project of Bachelors of Science (Computer Science). A Voice-Enabled-Urdu-Chatbot build using RASA specifically designed for farmers in Pakistan, they would be able to talk to Chatbot in Urdu Language and get to know the weather forecast for next week, Machinery prices, Fertilizers and Pesticides prices in Urdu Language.

📚 LangChain Deep Agents 课程式学习项目(deepagents + DeepSeek):minimal-deepagent 基础课(最小智能体→工具→多轮对话→记忆→多智能体→生产化)+ advanced-deepagent 进阶课(组件原理+中间件/Backend/编排/生产化)+ 生产级个人助手综合实战(SaaS 多租户+述职报告自动化)。每课一个可运行示例+中文注释。

Turn RAG system evolution into solution packages you can review, port - and reject. Two independently adoptable design packages (knowledge compilation / governed improvement), 40 claims with evidence coordinates, contracts, governance gates, D0-D5 handoff. Docs only: no runtime code, no unverified benefit numbers.

AI assistant for Dubai building-code compliance — hybrid RAG over 5 Dubai Municipality regulations. Next.js 15, Gemini 2.5 Flash, Supabase pgvector, 1202 passing tests.

Superpowers (obra/superpowers) as a DeepSeek Harness plugin: the methodology skills plus their session bootstrap

🫀 The autonomous heart of your GitHub repo — 24/7 AI agents that triage issues, review PRs, repair CI & keep docs fresh. Runs entirely in GitHub Actions. No server, no database. Provider-agnostic: Claude, OpenAI, Gemini & more.

SUGT - local AI gateway & skills console. Centralize LLM API keys, one OpenAI/Anthropic-compatible proxy for Claude Code, Codex & OpenCode. Homebrew & winget. Fully open source (Apache-2.0).

Unified fleet-wide agent memory system for DeepSeek Harness — shared Obsidian Vault for dsh, Codex, Claude Code & Hermes with dependency-free Python FTS5 core.

transformers safetensors qwen2 text-generation verilog conversational

Model Optimizer safetensors qwen3_next nvidia ModelOpt Qwen3CoderNext

Agent-native research workspace. A deterministic CLI that ingests arXiv / PDF / LaTeX / Zotero papers into an Obsidian-compatible knowledge graph, with local hybrid search (BM25 + vectors), claim-level provenance, and a nightly literature radar.

GroundLens is an execution verification runtime for AI systems and agents. It turns observable AI execution into deterministic, policy-governed evidence that can be independently verified.

SpyAI scans targets looking for ports, HTTP banners, and API paths typical of LLM agents, RAG, vector stores, and adjacent infrastructure. No authentication, no fuzzing, no exploits.

IssueClaw is a self-hosted AI coding assistant that lives entirely within your GitHub repository. By simply opening an Issue, you can chat with an AI, generate code, fix bugs, and automate tasks. It features persistent git-backed memory and runs entirely on GitHub Actions - meaning zero servers, zero databases, and zero monthly costs.

Structural code intelligence and persistent memory for AI coding agents, served over MCP (Claude Code, Codex).

One tool, ten AI models: GPT-4o, Claude Sonnet, Gemini Pro, Grok 2, Mistral, Llama 3, and more — all free.

Access Claude API (Opus and Sonnet) for free via key rotation pool. OpenAI-compatible wrapper.

Built as a free-time side project at Consultadd — in the spirit of Gmail's origin as a 20%-time project at Google. Not an official company product (yet), just one engineer's attempt to fix a problem every team has.

RAT: RunAnyThing via Fully Automated Environment Configuration (EMNLP 2026)

A UNIX-philosophy Agent application for macOS.

A lightweight, provider-agnostic agentic harness for VS Code. Now with free GLM 5.3 Flash

A transparent same-protocol LLM proxy that routes cheap tasks to Flash and hard ones to Pro by structural signals — no keyword guessing.

Vinci Code CLI — a terminal coding agent you can run with your own provider key (OpenAI, Anthropic, OpenRouter, and more). A distribution of Pi. MIT.

Execution-ready SEO + AI-SEO playbook you hand to an AI agent to build & rank sites for any niche — keyword research, crawlable build, on-page, technical, schema, pagespeed, Cloudflare deploy, AI-SEO (ChatGPT/Perplexity/AI Overviews), off-page, local, programmatic, measurement. SOPs + prompts + templates + checklists.

Curated, verified survey of the agent harness — the layer around an LLM (loop, tools, memory, context, runtime) that turns it into an agent. 116 papers + Claude Code / Codex / Cursor / OpenHands tools + Anthropic / OpenAI / Karpathy essays. Agent = Model + Harness.

Subscription window keeper that keeps AI coding plan windows active, locally or via a remote server.