Code agents

🧪🤖 Pytest plugin for OpenAI requests

Autonomous driving episode generation for the Carla simulator in a gym environment. This framework makes it easy to create driving scenarios to train/test the agent.

Decoding Claude Code's leaked 512K-line source — hidden features, 43 tool definitions, system prompts, architecture patterns

13 projects using ChatGPT API, Whisper, Embeddings, and DALL-E with Python.

A community-driven dictionary that simplifies software, engineering and tech terms for all levels.

A production-grade multi-agent system for comprehensive medical diagnosis and coding using specialized AI agents.

Python code to implement LLM4Teach, a policy distillation approach for teaching reinforcement learning agents with Large Language Model

Agent Skill. The only memory Claude Code needs exists in the file system.

A local dual-layer memory pattern for AI agents: a compact, human-readable markdown index paired with semantic retrieval from a local vector store, queried before each message. For cross-project recall where flat memory files or vector-only RAG fall short. Local-first. Reference implementation.

AI PM 실전 가이드 — Claude Code로 PM 워크플로우를 재설계하는 8파트 커리큘럼

Automated LLM Coding Tournaments. There can be only one (winning code solution from the competing AIs)

Free OpenAI-compatible API providing access to powerful AI models. No registration required, completely free to use. 🔥

AutoResearch + PromptFoo = AutoPrompter. Run it with Neo AI Engineer

Sync MCP servers, Skills, Agents & Commands across Claude Code, Cursor, OpenCode, Codex. One config, all tools.

A knowledge-graph AI writing assistant - tracks characters, threads, and world-building across your full manuscript, scaffolds your story bible from a premise and reviews chapters for consistency.

The IAM layer for AI agents: cryptographic identity, capability authorization, and audit trails for non-human identities. Open source.

CTX is the standard Cognitive Version Control System for AI. Structured working memory for AI agents. CTX preserves goals, tasks, evidence, decisions, conclusions, runbooks, and origins so agents can continue work instead of restarting from scratch.

Agent Knowledge Manager (akm). A CLI tool to manage AI agent knowledge, memories, skills, and more.

WormGPT is an experimental AI project designed strictly for educational, cybersecurity research, and defensive testing purposes only. It is intended to help developers, researchers, and students understand AI behavior, prompt injection risks, and security vulnerabilities in language models.

LangConfig is a open-source visual workflow builder designed to make AI agent development accessible to everyone. Build, test, and share LangChain and LangGraph agent workflows with an intuitive drag-and-drop interface.

Production-ready AI-powered CRM with 6 autonomous agents for lead qualification, email intelligence, sales pipeline, customer success, meeting scheduling, and analytics

Production-ready entrepreneurship skills for Claude Code — marketing, sales, operations, finance, and leadership. 24 skills built by a founder, for founders.

Turn a one-line idea into a branch with tested, reviewed, committed code. The brainstorm-to-commit pipeline for Claude Code.

Free, AI PR reviewer that runs entirely in GitHub Actions. No hosting required.

One of the early China-built lightweight Claude Code-inspired terminal coding agents, with autonomous execution, tool calling, custom agents, task control, and evidence-gated memory.

Python CLI tool to run queries against sqltatabases and convert the results to json, csv, excel in the command line or python program. Easy to use for Humans, automations, LLMS/Agents

一个更强调可控、可追踪和 Token 效率的 AI 多智能体开发工作流 || An AI multi-agent development workflow that emphasizes controllability, traceability, and token efficiency

Fail-closed execution firewall for AI agents: quarantine MCP tools, proxy OpenAI-compatible requests, emit signed receipts, and verify EvidencePacks offline.

A Claude Code / agent skill for building genuinely designed, editable Feishu / Lark (飞书) whiteboards — deliberate composition, real hierarchy, a gated pipeline with pre-render fit-check and independent design critique.

AI video production workflow platform for professional short drama teams. Timeline-first, harness-tested, agent-assisted.

Cheasee-Pi: Build Your Own Harness. Cheap. Easy. Secure. Token-saving agent harness with security guardrails and a Kanban git-oriented sub-agent framework.

Native macOS menu bar app that keeps your MacBook awake for AI coding agents - Claude Code, OpenAI Codex, Cursor - even with the lid closed. Open source (MIT), IOPMAssertion + scoped sudoers rule for password-free pmset, timed auto-off, login item, crash-safe self-heal. Built for overnight agent runs and clamshell mode without an external display.

Agentic Coding platform for Professional SWE work focusing on output quality. Self-review, supervision, dynamic workflows, dynamic agent loops.

Self-hosted gateway for Claude Code, Codex, Antigravity, Grok Build, OpenCode. Run AI coding agents on your own infra with a shared local-first memory layer. Drive from web, mobile, Telegram / Slack / Discord / Feishu / DingTalk / WeCom. Open REST + WebSocket API. Apache 2.0.

Nirvana-OS — source-available engine for agentic work (squads, businesses, mind-clones). Developed in the open. Runtime-agnostic: Claude Code, Codex, Gemini-CLI, Antigravity, Hermes.

The fastest way to run DeepSeek, OpenAI, NVIDIA, Groq, OpenRouter, Moonshot/Kimi, Cerebras, and Ollama models in GitHub Copilot in Visual Studio and Visual Studio Code

PlantUML diagrams (.puml) from natural language via Kroki — no Java install. 10+ UML types, 5 themes, PNG/SVG, multi-agent.

safetensors gguf qwen3 function-calling tool-calling codex

A traceable personal memory layer that carries verified state across apps, models, and AI agents.

macOS & Windows Codex Desktop 梦幻换肤 Skill｜由世事宜AI免费提供

Local web workspace for coding agents — Claude, Codex, Gemini, and more in one inspectable, LAN-ready hub.

MCP server for Z-Library: search, download, and process books into RAG-ready text for AI assistants (Claude Code, Claude Desktop, Cline, RooCode)

Faster runtime for coding agents. Make coding agents 25% faster and 30% cheaper on average while keeping the quality same or more. Same Task, Same Quality, Faster and Cheaper.

OpenAI-compatible proxy for Microsoft 365 Copilot — use the Copilot licence you already have as a backend for pi, openclaw and other coding agents. Prompt-emulated tool calling, no extra licence, no MCP.

Give Claude Code a persistent memory — it remembers you, your projects, and your decisions as plain Markdown files you own. Fully local. Just say `start`.

Semantic retrieval for AI agents: high-recall memory and code search that cuts coding-agent token usage by nearly 50%

Multi-frontend agent relay: run Claude Code, Codex, local, and AG-UI agents from Discord or Microsoft Teams

Open Swiss legal corpus + MCP server: 1M+ court decisions (1875–today), 21k laws, 10M-edge citation graph, 42 MCP tools. CC0 data, MIT code. Live at mcp.opencaselaw.ch

Native macOS agent control for Codex and Claude — monitor, approve, ask, and jump back from the menu bar.

把Claude Code的UltraCode模式带给DSH，把 DSH 的一次性多 Agent 调度，升级为可生成、可保存、可治理、可观察、可恢复的 Workflow 层

The FunctionChain is a tool that simplifies and organizes the process of invoking OpenAI functions in your Node.js applications. With this toolkit, you can easily scaffold out and isolate all the OpenAI function calls you need, making your code more modular, maintainable, and scalable.

Build low-code, event-driven AI applications and workflows using **any** OpenAI API-compatible framework.

A fast, agnostic, and powerful Go AI framework for one-shot workflows, building autonomous agents, and working with LLM providers.

A Web app stack written in FastAPI, Qdrant, and React for creating AI projects

Chatbot based on Almanac of Naval Ravikant. Uses OpenAI's chat completion API

A flexible PHP library providing access to various AI models (Gemini, OpenAI, Anthropic, DeepSeek, Grok) via a consistent interface.

A collection of cookbooks to help developers get started quickly with the Firecrawl API.

Twitch Streamer GPT is a NodeJS-based Twitch enhancement tool, offering interactive stream experiences with AI-powered automated responses, voice command activations, and advanced modules. It's easy to set up and suited for users of all tech levels.

Get the information of a Github Repository using the power of LLM.

Text to p5.js Generative Editor powered by OpenAI GPT

A basic AI chat using the OpenAI API and its GPT-3 models

A SwiftUI app demonstrating how ChatGPT interacts with DALL·E and HuggingFace models for iOS and macOS.

Set of instructions for your favorite AI agent (e.g. Cursor, Copilot, Claude, etc) for a "DHH-driven" Ruby on Rails development philosophy.

Dcup - Advanced RAG for Personal Knowledge ☕

🧙‍♂️ Text2SQL 是基于 SpringBoot、MySQL、Vue2 以及 ChatGPT，通过用户输入的表结构和需求，神奇地生成具体的 SQL 语句！用户可以导入数据库和表结构，选择具体的表，生成独特的 SQL 语句。致力于解放开发同学的双手，扎根于更深层次的业务中。Text2SQL 更适合对 SQL 感兴趣的开发同学呦 🌟

Unofficial Anthropic API client for Elixir. Integrate Claude, Anthropic's powerful language model, into your applications.

Deep Reinforcement Learning (DRL) algorithms have been successfully applied to a range of challenging simulated continuous control single agent tasks. These methods have further been extended to multiagent domains in cooperative, competitive or mixed environments. This paper primarily focuses on multiagent cooperative settings which can be modeled for several real world problems such as coordination of autonomous vehicles and warehouse robots. However, these systems suffer from several challenges such as, structural credit assignment and partial observability. In this paper, we propose Recurrent Multiagent Deep Deterministic Policy Gradient (RMADDPG) algorithm which extends Multiagent Deep Determinisitic Policy Gradient algorithm - MADDPG \cite{lowe2017multi} by using a recurrent neural network for the actor policy. This helps to address partial observability by maintaining a sequence of past observations which networks learn to preserve in order to solve the POMDP. In addition, we use reward shaping through difference rewards to address structural credit assignment in a partially observed environment. We evaluate the performance of MADDPG and R-MADDPG with and without reward shaping in a Multiagent Particle Environment. We further show that reward shaped RMADDPG outperforms the baseline algorithm MADDPG in a partially observable environmental setting.

Makes your codebase agent-friendly and keeps it that way. Small files so agents don't waste tokens. CLAUDE.md in every directory so they don't have to guess. Tests so they check their own work. Run once, it bolts rules into your global config — every project enforces standards automatically.

Neuro-Symbolic-Causal AI - Project Chimera | 🌌 An open research project exploring formal verification of AI agent decisions, combining symbolic reasoning, causal inference, and runtime policy enforcement.

Claude Code Agent Team GUI Monitor - Tauri v2 + React + TypeScript macOS desktop app with pixel art office view

An agentic AI for tech career coaching 程序员技术学习AI智能体 (based on JoyAgent-JDGenie with local RAG and code analyzing)

Write your repo rules in Markdown. AI enforces them — triages issues, reviews PRs, fixes bugs, merges code, cuts releases.

AI-powered architecture analysis plugin for Claude Code. Generates diagrams, detects anti-patterns, and scores your codebase 0-100.

Prompt Hardener analyzes prompt-injection-originated risk in LLM-based agents and applications.

🤖 Local AI agent for your laptop. Voice activation, multi-step tool use, 7-layer memory, human-in-the-loop approvals. Zero cloud. Zero API keys. Zero telemetry.

This repository is a paper summary of the latest progress in cooperative/collaborative/multi-agent perception datasets in autonomous driving scenarios such as V2V/V2I/V2X/I2I/Roadside Perception.

🐍 Асинхронный Python-фреймворк для разработки ботов MAX Bot API (max.ru)

Multi-agent system that designs and generates complete agent team structures through conversational interviews. Supports Claude Code and Codex.

Turn any repo into a governed AI workspace. Quality gates, security scanning, and risk management — enforced locally via git hooks. Works with Claude Code, GitHub Copilot, Cursor, Gemini & Codex.

iMessage integration for Claude Code - autonomous agent via iMessage

RJ Auto Metadata is a desktop application that uses AI to automatically generate titles, descriptions, and keywords for your media files.

AI agent set for cloud security purple teaming, runs inside Claude Code, Gemini CLI, and Codex.

An autonomous agent that acts as a DEF CON-level Certified Ethical Hacker, using tools such as Nuclei, sqlmap, ffuf, Burp, ZAP, and the Social Engineering Toolkit (SET), (labs, CTFs, HackTheBox, TryHackMe, and bug bounties with a written scope).

Deterministic, zero-dependency Python firewall for AI agents — MCP rug-pull, memory poisoning, indirect injection, exfil channels. 44 compliance templates (US/CN/JP/EU).

Bring your own key (BYOK) relay to enable lightweight frontend-only apps to use LLMs using the user's API keys without hitting CORS.

Claude Code hook that detects context compaction and injects a reminder to re-read AGENTS.md, preventing post-compaction rule amnesia in long sessions

Native C++ AI assistant module for the Godot 4.7 editor - multi-provider LLM chat, GDScript code generation & execution, scene-aware context, on-demand API docs and gotchas injection

Deep code indexing MCP server for AI agents. 25 tools: hybrid FTS5 + embedding search, call graphs, git blame/hotspots, build system analysis. Multi-repo workspaces, GPU-accelerated semantic search, 10 languages via tree-sitter. Fully local, zero cloud dependencies.

A targeted resource for mastering LangChain, featuring practice problems, code examples, and interview-focused concepts for building AI applications with Python. Covers chaining LLMs, memory management, and tool integration for technical interview success.

A collection of examples demonstrating how to build AI agents that integrate with the Xero API using various agentic frameworks and the Xero MCP (Model Context Protocol) Server.

LLM orchestration toolkit for agent workflows: planner + workers + synthesis, optional router (LLM + learned fallback), supports OpenAI/Anthropic/Ollama/llama.cpp, real scraping with caching, MCP server integration, and a TUI chat UI.

面向 Claude Code、Codex、Open Code等Ai Agent 的确定性公文排版 Skill，可直接渲染为规范公文格式的 .docx（字体、字号、行距、页码等）。

The open-source marketing agent workspace for Claude Code. 42 commands, 12 hooks, 4 industry starters.

Airtap skill for agents like codex, claude, openclaw

Free-tier LLM API key pool with rotation, cooldown handling, and an OpenAI-compatible proxy. Use with Hermes Agent or any OpenAI-compatible tool - no paid API key required.

Bricks is the open-source, fully local alternative to Clay.com. It combines AI agents, advanced web scraping, and Puppeteer automation to turn raw CSV data into actionable business intelligence — without subscriptions, vendor lock-in, or sending your data to the cloud.

Local OpenCode-backed LLM gateway for OpenAI, Anthropic, Gemini, and Responses API-compatible tools, with streaming and tool/function calling.

金蝶云星空 K3Cloud MCP Server，让 AI 助手（Claude Desktop、Claude Code、Cursor、Cline、Cherry Studio、Openclaw 等任意支持 MCP 协议的客户端）通过自然语言查询和操作金蝶 ERP 系统。

Anthropic API reverse proxy with prompt-cache injection and request body transforms