Code agents

Zero-LLM agent memory for Claude Code and AI agents: local-first BM25, dense-vector, and reciprocal-rank-fusion retrieval. Returns original passages verbatim by default. Available on PyPI as fidelis-memory. MIT.

Mikrotik RouterOS dashboard with built-in AI assistant, analytics and network map. Supports Claude, LM-Studio and Cloudflare AI Agent models.

Connect Chromium to LLM agents via token-efficient DOM compression. 50-200 tokens per page.

Autonomous multi-agent coding system on local LLMs. No frameworks, no API costs. Plan → Build → Test → Fix with Qwen 80B + 7B on your own GPUs.

A flexible platform that provides Docker & Kubernetes backends, a lightweight CLI (mcpt), and client utilities for seamless MCP integration. Spin up servers from templates, route requests through a single endpoint with load balancing, and support both deployed (HTTP) and local (stdio) transports — all with sensible defaults and YAML-based configs.

An intelligent customer support system powered by LangGraph and LangChain that uses Retrieval-Augmented Generation (RAG) to provide accurate, context-aware responses to customer queries. Built with FastAPI, FAISS, and multi-stage validation for production-ready deployment.

Benchmarking platform for evaluating AI agents on CTF-style tasks and tool-use workflows.

A transparent proxy system for inspecting, debugging, and modifying any traffic from the agent container, focused on LLM-based coding agents. Works with Claude Code, Codex CLI, Cursor, Aider, and any other tool that calls LLM APIs.

OpenArchFlow is an open-source Progressive Web App designed for cloud architects, DevOps engineers, and developers. Generate interactive AWS architecture diagrams from natural language descriptions using AI, Large Language Models (LLMs), AI Agents, and AWS MCP (Model Context Protocol).

Low-Cost Cross-Domain Web Structured Information Extraction using specialized LoRA adapters.

Local-first web UI for exploring your X/Twitter bookmarks

A privacy-focused web UI for LLMs (OpenAI-compatible APIs and Ollama) designed to run directly from a portable directory on your filesystem.

A curated directory of every first-party skill, capability, and tool across Claude, ChatGPT, Gemini, Codex & DeepSeek — searchable, filterable, deep-linkable.

Free & Open Source AI Agent — Self-Hosted, Self-Learning, No Vendor API. Qwen2.5-7B + LoRA. 35 tools. Built on Islamic Epistemology (IHOS).

AI-powered development orchestrator: Supports Claude Code, Codex and Langchain Deepagents + Openrouter backend

Manage and orchestrate AI coding agents (Claude Code, Copilot, Goose, etc.) on remote machines — task decomposition, multi-agent coordination, and real-time visibility across environments.

🏭 Agentic dark factory for AI building with sealed-envelope testing

Agent Droid Bridge gives AI agents programmatic control over Android devices and emulators via ADB, exposed as an MCP server.

Meridian - A graph-powered AI chat application integrating intelligent parallelization for advanced, aggregated conversational experiences. Nuxt3 & Python.

Perfetto trace analysis via MCP — PerfettoSQL queries plus dedicated Chrome tools for scroll jank, page loads, and main-thread hotspots

Know what you spend, see what you send. Lightweight local proxy that logs every LLM call - costs, tokens, full prompts and responses.

Runtime guardrails for Claude Code. Auto-approve what's safe, gate what's risky, block what's dangerous. Dual enforcement, full audit trail. MIT.

Unofficial Node.js client for gemini.google.com.

Supercharge Claude Code with 109 specialized components — 22 agents, 32 commands, 55 skills across 13 modular plugins. Deep reasoning, autonomous dev loops, DDD architecture, design system enforcement, git automation, and more.

Apple-native iOS/macOS app for document intelligence, OCR, cited answers, and source-backed retrieval over PDFs, scans, and user-controlled files.

Self-hosted API gateway for Claude Code on Amazon Bedrock — unlock extended thinking, web search & tool use with team budgets, OIDC SSO, and an admin portal

Real-time Claude API usage monitor on ESP32-S3 with an LCARS Star Trek display — tracks costs, tokens, rate limits, and Claude.ai subscription status

Hybrid Knowledge Graph + RAG security intelligence platform for CVE analytics and semantic search. Llama3 integrated.

An end-to-end RAG chatbot using FastAPI, LangChain, ChromaDB & Streamlit with Multi-LLM support to answer questions from uploaded PDFs.

Persistent, queryable memory for AI agents with semantic recall, layered storage, snapshots, and multi-agent scoping

Local-first Memory Framework for AI Agents · 99.2% LongMemEval-S retrieval @ k=10 · Supports Claude · Antigravity · LangChain · Hermes · Gemini · OpenCode · OpenClaw · MCP-native and plugins · Hybrid search (FTS5 + vector + MMR) · GDPR · FIPS 140-3 ready · 100% local (fully offline) or cloud capable

Self-learning context infrastructure for AI engineering agents. Connect repos, docs, and tools into one org-scoped knowledge graph and expose it through a single MCP.

Flutter clean architecture skill for AI agents. Riverpod 3.x codegen, Freezed 3.x sealed classes, and GoRouter patterns.

context engineering for agentic coding

An "all-you-need-to-know" practical guide to using AI and LLMs responsibly and effectively for economics and finance research: from core skills and ChatGPT/Claude Projects to coding, writing, reproducibility, and automation. 小红书账号：SuperJay

A full-featued Calibre MCP server with granular permissions for read/write/import/etc. Locally tested and freshly released, consider this an alpha. Back your library up before you give an agent write permissions.

🔍 一键检测 AI API 服务商是否如实提供所标注的模型 | 支持 OpenAI / Claude / Gemini / DeepSeek 等 9 大模型原厂对比验证

RadCrew.org website with LLM-powered chatbot

AI dependency migration analyzer — reads every changelog so you don't have to.

OpenAI Codex CLI — real-SIM SMS verification via VirtualSMS MCP. Drop-in .codex/config.toml, 18 tools, 2000+ services, 145+ countries.

Chinese traditional fortune-telling AI skill (BaZi 八字 + Zi Wei Dou Shu 紫微斗数) for Claude Code, Codex, Cursor, Trae and any agent platform. PolyForm Noncommercial.

bb-huge 🤗 , Personal bug bounty findings hub and bug bounty orchestration for multiple agents

Agentic Testing Harness

Cross-IDE, multi-stack agent skill for OpenUI (the Open Standard for Generative UI). Adds OpenUI to existing projects across 12 backend stacks, any LLM provider, and 11 agent platforms. Scaffold, integrate, validate.

Team memory sharing for claude-mem — sync AI memories across developers, with Claude Code plugin, github action and knowledge distillation

Claude Code skills to bootstrap, compact, and maintain AI-readable project documentation — doc-init · doc-compact · doc-update

Run your company with AI agents. Open source harness for operating a full company with specialized Claude Code squads.

Let Codex test multiple fixes in isolated worktrees, compare evidence, and apply only the safest proven solution.让 Codex 不再盲改代码，而是并行测试多个方案，用证据选出最稳的修复。

Auto-Browse: AI Enabled Browser Automation

Fragaria is a powerful and flexible Chain of Thought (CoT) reasoning API that leverages various Language Model (LLM) providers and incorporates Reinforcement Learning (RL) techniques to solve complex problems and answer intricate questions.

Multi-provider Android AI chat SDK with Jetpack Compose UI, headless Kotlin API, Room history, and support for OpenAI, Gemini, Claude, and Grok.

A tiny macOS desktop widget for your Claude Code 5-hour usage window: live reset countdown and % used. Matches claude.ai.

Visual RAG workflow studio that auto-generates eval sets, runs RAGAS evals, and deploys workflows as local HTTP APIs.

One command, 28+ top AI CLIs — an interactive launcher for terminal AI coding agents (Claude Code, Codex, Gemini, Aider…). Uzbek & English

Parallel coding agent workspace with native terminals and git worktrees.

Modular, agentic framework and MCP platform you self-host. Build and run your own AI agents, connect Claude Code, Claude Desktop, or Cursor as a native MCP connector, self-host your apps, and share over encrypted P2P — on hardware you own, with local or cloud models. Open source.

Autonomous AI research harness for Claude Code. Sibling to rockie-codex (OpenAI Codex CLI). Apache-2.0.

AI-DLC (AI-Driven Development Life Cycle) claude code plugin

GraphRAG over any codebase — a Neo4j knowledge graph of code entities plus a 3072-dim vector index, bridged so one natural-language query walks structure and ranks by meaning. Python · Neo4j · OpenAI embeddings · Streamlit · Docker. Hybrid semantic + structural code search.

2-in-1 for Claude Code: a live system-tray usage widget (5h/7d plan limits) + local token & cost analytics — per project/model/tool, budgets, forecasts. Unofficial; not affiliated with Anthropic.

Публичный пакет инженерных skills для Codex

Leave agents running without giving up control.

基於 Google ADK 與 MCP Toolbox 的保險科技推薦 Agent 原型。核心 Agent 能理解保險需求、主動追問，並透過受控 SQL 工具查詢醫療、意外、家庭與收入保障產品。後端採 FastAPI 提供 SSE 串流與 WebSocket 即時互動，前端以 Next.js 呈現對話介面。知識庫結合 PostgreSQL 與 pgvector 實現 FAQ 語意搜尋與 RAG 查詢，並支援多模態 Live Agent 語音互動。安全層涵蓋 JWT 驗證、PII 脫敏與 SHA-256 雜湊鏈防篡改稽核日誌。品質保證採 ADK evalsets 評估驅動開發，部署支援 Docker Compose、Cloud Run 與 Terraform 多環境配置。

Local-first dashboards your AI agent builds and keeps maintaining, over data that never leaves your machine.

Hermes Agent profile for adaptive multi-agent coding teams with GitHub issue-to-PR automation, blind critique, synthesis, CI schemas, and fail-closed unattended merge policy.

🔊 Text-to-Speech MCP plugin for Claude Code - hear audio feedback while coding (OpenAI TTS)

Autonomous IT service desk built entirely on Microsoft 365 — Copilot Studio, Power Automate, and SharePoint. Ticket → triage → approve → execute, no per-seat SaaS fees. Microsoft Agent Academy Hackathon winner.

Command-line client for the Harmont CI platform

Durable, local-first memory for AI coding agents over MCP — zero-dependency, curated & semantically de-duped, you own the data (SQLite + Markdown). Works with Claude Code, Codex & any MCP host.

Turn any OpenAPI spec or Postman collection into an ownable Python MCP server in 30 seconds

Production-grade framework for building multi-agent AI systems. Graph-based orchestration, LLM-agnostic (OpenAI, Google GenAI, Anthropic), 3-layer memory (Redis cache + Postgres + vector store), live agents, parallel tool execution, and native MCP. Ships a full ecosystem: backend, REST API + CLI, TypeScript SDK, and React playground

A powerful offline chat application that provides AI mentors specialized in various tech domains. Learn programming with intelligent, emotionally-aware AI mentors that understand your needs and guide you through your development journey.

Local MCP server that helps coding agents stop re-reading the same code every turn. Repository intelligence, cache-write optimization, and long-term session memory. 100% local. No API keys.

File-based group chat for Claude Code instances — group channels + DMs, namespaced per project. Installable as a skill or plugin.

A working repository for designers who use Claude as a collaborator. Skills, principles, project templates, and a career vault.

TwinCAT coding agent for Beckhoff PLC development, with strict rules for Structured Text, naming conventions, PLC-aware tooling, syntax checks, and safe project validation workflows.

Self-hosted AI API gateway 鈥?proxy OpenAI / Claude / DeepSeek behind one endpoint; works with Codex CLI / OpenCode / Cherry Studio.

A lossless token-optimization codec for LLM agents — cut input tokens up to 82%, nothing dropped, every fact kept. Drop-in proxy + one-line Python API for Claude Code, Codex, Cursor, OpenAI, Anthropic & Ollama. pip install foveance

Spec-Driven Development skill pack for the Pi coding agent: steering docs, .status gates, EARS, and a PRD→SPEC→TASKS→EXEC→REVIEW pipeline.

transformers pytorch llama text-generation code math

transformers safetensors mistral text-generation openchat C-RLFT

transformers safetensors qwen2 text-generation code codeqwen

gguf imatrix conversational qwen3_next ik_llama.cpp text-generation

transformers pytorch codegen text-generation arxiv:2203.13474 license:bsd-3-clause

mlx safetensors gemma4_assistant speculative-decoding mtp gemma

transformers safetensors olmo3 text-generation conversational en

Python implementation of coding agent core mechanisms, with architecture comparison across Claude Code / Kimi Code / Codex

Local AI coding-agent history + API cost analytics (loopback-only).

AI-powered autonomous trading agents for Polymarket prediction markets. This framework enables developers to build sophisticated trading bots using state-of-the-art LLMs, RAG (Retrieval-Augmented Generation), and superforecasting methodologies. Built with Python 3.12+, LangChain, FastAPI, and modern development tooling.

Debugger CLI for AI agents — run, control, and inspect binaries programmatically with structured output for LLM tooling

Local dashboard to run parallel Claude Code & Cursor agents — kanban fleet view, one permission inbox, in-app diff review. macOS, MIT.

A self-documenting, self-extending multi-agent system. Topics are miniature GOOP instances. Markdown is the API between human and machine. Built on Hermes Agent by Nous Research.

Self-hosted, Elixir-native alternative to LiteLLM: one API for multi-provider routing, fallbacks, budgets, keys, and observability.

Turn Claude Code into an assistant with persistent memory: one interview-driven prompt builds your context files, rules and decision log, and every session starts already knowing you.

Claude Code plugin: Fable 5 as a token-frugal orchestrator with tiered Opus/Sonnet/Haiku agents

Evidence-grounded AI agent for Java code repair with LLM-guided patches and human-in-the-loop review.

Log Claude Code sessions to Opik, the open-source LLM observability and evaluation platform, built by Comet. Tracing, evaluation, and skills for observable AI applications.

한국 취업 자기소개서를 기업조사→초안→글자수→리뷰→윤문→검증 8단계로 완성하는 Claude Code 플러그인

Give Claude a checklist before bed. It can't clock out until every box is ticked, never stalls on a question, and revives itself if the API dies at 3 AM.

Don't write a bug report — record it. Local MCP: screen recording → transcript, frames, OCR, wall-clock evidence for coding agents.