Explore

Python MCP server that indexes repositories with tree-sitter and provides coding agents precise symbol-level source retrieval and structural queries. Integrates with clients such as Claude Code, Cursor, and VS Code to reduce code-exploration context use.

MCP | Python · MCP server · code retrieval

Xianyu/Goofish smart customer-service system: FastAPI + SQLite + Playwright backend that drives AI auto-replies, keyword rules and auto-delivery across multiple marketplace accounts, with a full web admin console and Docker deployment.

AI Tools | Python · auto-reply · customer-service

A Python starter template for production-oriented AI agent APIs using FastAPI and LangGraph. Includes checkpointing, tool calling, mem0 long-term memory with PostgreSQL/pgvector, LLM fallbacks, authentication, tracing, metrics, and evaluation utilities.

AI Agents | Python · LangGraph · FastAPI

Godot AI is a production-grade MCP server and editor plugin that connects LLM clients like Claude Code or Codex to a live Godot 4 editor, exposing 46 tools and 120+ operations to build scenes, scripts, signals and UI by prompt.

MCP | GDScript · godot · mcp-server

MCP server and CLI that connect AI assistants to South Korea’s official legal information APIs. Search statutes, precedents, and ordinances; verify citations; and run date-specific legal research and analysis.

MCP | TypeScript · MCP server · Korean law

Vibe-Research is a local, AI-agent-driven personal investment research workbench for A-share, US and HK markets. It runs six-stage stock research, news radar, bull/bear debate and backtesting through your own Codex/Claude/API model, with a controlled MCP toolchain, deterministic calc engine and evidence-backed reports.

AI Agents | TypeScript · ai-agent · financial-research

A local-first desktop and browser agent with a CLI and Tauri sidecar. It runs tool-using workflows with local llama.cpp models or cloud models, supports MCP tools, and stores sessions and memory on the user's machine.

AI Agents | TypeScript · local-first · computer-use

A Chinese-language, 16-chapter hands-on course for building production-oriented agents with LangChain, LangGraph, and Deep Agents. Covers planning, subagents, skills, memory, MCP, sandboxes, permissions, and orchestration.

AI Agents | Astro · Deep Agents · tutorial

Docs compiler that turns one Markdown source into a static site plus machine-readable outputs: llms.txt, Open Knowledge Format bundles, an MCP server for agents, and a RAG AI assistant. TypeScript CLI, MIT licensed, installable via npx or Docker.

AI Tools | TypeScript · documentation · static-site-generator

MCPHub is a self-hosted MCP gateway and control plane: connect local and remote MCP servers once, expose them via stable MCP endpoints with per-user credentials, OAuth 2.0, smart vector routing, health checks and a web dashboard. Ships as Docker/npm with optional PostgreSQL mode.

MCP | TypeScript · mcp-gateway · mcp-router

A pack of 20 Claude Code agent skills (markdown workflows) covering resume/ATS optimization, job-fit analysis, cover letters, interview prep, salary negotiation, and LinkedIn tuning. Installable via npx skills add into Claude Code, Cursor, Windsurf and other agents.

Skills | claude-code · agent-skills

A self-hosted framework for building industry news sites. It collects from configurable sources, uses LLMs to filter and score articles, writes summaries, clusters related coverage into events, and publishes ranked topics and briefings.

AI Frameworks | TypeScript · AI news aggregation · LLM content selection

An agent skill for designing and implementing interactive web animations from generated or existing image and video assets. It provides workflows and scripts for frame cleanup, compression, resource selection, and runtime control by scrolling, pointer movement, dragging, touch, or state changes.

Skills | Python · agent skill · web animation

An MCP server and FreeCAD addon that let compatible AI clients create and edit CAD models, inspect documents, execute Python scripts, and run FEM analyses.

MCP | Python · FreeCAD · CAD

MTPLX is an Apple Silicon LLM inference engine and Mac app that uses Qwen's native multi-token prediction heads for exact speculative decoding (1.6x-2.24x faster decode at any temperature). It serves local models over OpenAI- and Anthropic-compatible APIs for coding agents and chat.

AI Frameworks | Python · apple-silicon · mlx

Lanhu MCP server (FastMCP + Playwright) that exposes blue-lake/Lanhu Axure requirement docs and UI design specs to any MCP client, extracting text, annotations, screenshots and design assets, plus generating HTML+CSS references and a shared team knowledge board. Requires a vision-capable model and a Lanhu cookie.

MCP | Python · mcp-server · fastmcp

Hosted remote MCP server exposing 250+ tools for Google Ads, Meta Ads, GA4, and Search Console to AI clients such as Claude, ChatGPT, Cursor, and n8n. OAuth login handles account access, and changes require approval.

MCP | remote-mcp · advertising

An MCP server that gives AI clients read and write access to Figma, including design-token extraction and sync, component and layout creation, accessibility audits, and FigJam and Slides tools. Runs locally with a Figma Desktop bridge or remotely on Cloudflare Workers.

MCP | TypeScript · Figma · design systems

A Claude Code plugin / agent skill pack that teaches coding agents to build with TypeSafe's System One API for typed decisions and probabilities. Installs via the Claude plugin marketplace or skills.sh and helps design workflows, route uncertain cases to humans, and write typed judgments in code.

Skills | agent-skills · claude-code-plugin

Self-hosted security operations platform that ingests telemetry, detects and correlates threats, and uses auditable AI agents to triage alerts and investigate incidents. Includes local Ollama inference and an MCP server; response actions require human approval.

AI Agents | Python · AI security · SOC

An 11-skill workflow for creating AI short dramas and comics, from source analysis and episodic scripts to storyboards, image/video prompts, reviewed production tasks, and editing. Installs into Claude Code, Codex, or other Agent Skills-compatible environments.

Skills | Python · agent-skills · AI screenwriting

Provider-neutral Agent Skill (SKILL.md + reference docs) for designing, auditing, and refactoring agentic harnesses: loops, tool permissions, context/memory compaction, evals, cost and security. Installs into Codex or Claude Code skills directories.

Skills | agent-skill · claude-code

Zero-dependency C++23 CLI plus MCP server that builds a deterministic call graph for coding agents, answering callers, uses, impact, tests-to-run and diff-review queries in a fraction of the tokens of grep-and-read. Also installs task-shaped agent skills for Claude Code, Codex, Cursor and others.

AI Coding | C++ · mcp-server · code-navigation

A notebook-based guide to building a LangGraph email agent, covering core agent concepts, evaluation, human review, and persistent memory. Includes Gmail API integration and deployment guidance.

AI Agents | Jupyter Notebook · email assistant · human-in-the-loop

LangGraph-based multi-agent tool that turns a company name into a structured, source-backed research briefing using Tavily search plus Gemini and OpenAI models, with a FastAPI backend, React UI, streaming progress, and PDF export.

AI Agents | Python · multi-agent · research-agent

NVIDIA NeMo Data Designer is a Python framework for generating and augmenting synthetic datasets with LLM, sampler, and image columns. It supports multimodal seeds, MCP tool-use traces, validators/LLM judges, plugins, and scalable resumable runs.

AI Tools | Python · synthetic-data · data-generation

TypeScript CLI + MCP server that parses Korean documents (HWP 3.x/5.x, HWPX, HWPML, PDF, XLS/X, DOCX, images) into Markdown, with lossless HWPX roundtrip, form filling, document diff, RAG chunking, OCR, and PII redaction. Registers 17 document tools for Claude Desktop, Cursor, and Codex.

MCP | TypeScript · hwp · hwpx

Memanto is a Python memory agent that runs alongside your AI agents to curate, consolidate, reconcile and expire their long-term memory, with point-in-time recall, policy-based forgetting and portable OKF export. Ships as a CLI plus local REST API and dashboard, and can run fully on-prem with Docker and Ollama.

AI Agents | Python · agent-memory · long-term-memory

LLM-powered language-learning web app: curate vocabulary from books or videos, then practice via 5 generated quiz types, reading passages, image-based word discovery, and an OpenAI voice tutor that can edit your lists. React/Vite frontend, Node backend, MongoDB, Docker.

AI Tools | TypeScript · language-learning · vocabulary

A skill pack that adds a structured two-phase deep research workflow (outline, then parallel web investigation) to Claude Code, OpenCode, and Codex, with human-in-the-loop control and markdown report output. Install via copied skill/agent files plus pyyaml.

Skills | Python · claude-code · codex

ArcKit provides architecture-governance commands, skills, and specialized agents for AI coding assistants. It supports workflows such as requirements, risk and business cases, technology research, diagrams, vendor selection, and design reviews across multiple assistant platforms.

Skills | JavaScript · enterprise-architecture · architecture-governance

Self-hosted Apache-2.0 gateway that exposes Codex, Claude Code, Hermes, PI and DSH harnesses through one OpenAI Responses-compatible API, with sessions, streaming, files, cancellation and failure handling under the Unified Harness Protocol.

AI Agents | Python · agent-harness · unified-api

ToolHive is an enterprise-grade platform for running and managing MCP servers. It wraps each server in an isolated container, adds a gateway, registry, and Kubernetes operator, and enforces per-request identity, authz, and observability so teams can self-host MCP in production.

MCP | Go · mcp-servers · mcp-security

A testing and evaluation platform for MCP server developers. Inspect and debug servers, test MCP and ChatGPT apps, trace protocol and OAuth activity, and run evaluations or conformance checks locally and in CI.

MCP | TypeScript · server inspector · debugging

A Claude Code skill that downloads YouTube videos, uses AI to generate semantic chapters, clips segments with FFmpeg, translates subtitles into bilingual EN/ZH SRT, burns them in, and writes social-media summaries. Installs via npx skills or install_as_skill.sh.

Skills | Python · claude-code · claude-skill

Plugin that gives coding agents self-managing long-term memory and unbounded context: it captures durable project knowledge, consolidates it in the background, and auto-recalls it each turn across OpenCode, Pi, and OMP. Includes a CLI setup/doctor wizard and shared cross-harness memory store.

AI Agents | TypeScript · agent-memory · context-management

LiveAgent is a local-first AI agent desktop client built with Tauri 2, React 19 and Rust. It runs multi-model agents (Claude/Codex/Gemini) that operate the local file system and Bash, bridges MCP servers, loads Skills packages, and can be remotely controlled from a browser through its Go Gateway.

AI Agents | TypeScript · desktop-client · local-first

Pinvou Agent is a Tauri/Rust desktop AI agent workspace combining work tasks, prompt-to-design generation, and ACP coding agents (Codex, Claude Code, Kimi) in one app. It adds local knowledge base RAG, MCP/CLI connectors, Skills, and workflows, and runs fully local with vLLM or any OpenAI-compatible endpoint.

AI Agents | Rust · desktop-agent · tauri

Open-source Computer Use service exposed as an MCP server, letting any agent or MCP client control desktop apps on macOS, Linux, and Windows via accessibility APIs. Ships CLI installers for Codex, Claude Code, Gemini CLI, and opencode plus an agent skill.

MCP | Swift · computer-use · gui-automation

A native Zig plugin that exposes x64dbg debugging features as MCP tools over authenticated HTTP. AI assistants can control execution, inspect and modify process memory, manage breakpoints, and analyze binaries.

MCP | Zig · MCP server · x64dbg

MCP server exposing 97 browser-automation tools (or a 20-tool core) on real Chrome via nodriver and CDP, focused on evading anti-bot checks plus element cloning and network interception. Install locally with a trusted MCP client such as Claude Code or Cursor.

MCP | Python · mcp-server · browser-automation

A Go-based MCP server that gives AI clients access to Kubernetes and OpenShift clusters. It supports resource and pod operations, Helm releases, Tekton pipelines, multi-cluster access, and configurable access controls.

MCP | Go · Kubernetes · OpenShift

Orkas is a local-first desktop app where a Commander LLM plans goals and dispatches nine built-in specialist agents (research, writing, slides, office docs, video, image, UI, SEO, dev). It also drives local CLI agents like Claude Code, Codex and OpenCode, with per-agent skills, memory and reflection-based self-improvem

AI Agents | TypeScript · multi-agent · agent-orchestration

A Claude Code and Codex skill that guides agents through researching a topic, writing narration, generating TTS and storyboards, and building a Remotion explainer video with subtitles and quality checks. Includes a reusable TypeScript template and supports Chinese and English.

Skills | TypeScript · agent-skill · AI-video

A full-stack TypeScript and React framework for building type-safe MCP servers and interactive MCP Apps for ChatGPT, Claude, and compatible clients. Includes a local emulator, hot reload, CLI, and agent-oriented development tools.

MCP | TypeScript · MCP Apps · MCP servers

A harness for Claude Code and OpenAI Codex that adds persistent project context, structured development workflows, quality gates, review, and runtime verification. Includes an installer and a console for human oversight.

AI Coding | JavaScript · coding-agents · context-engineering

LetsFG provides an MCP server, CLI, and Python/JavaScript SDKs for AI agents to search and book flights and hotels. It returns live offers and supports booking workflows that can produce airline PNRs.

MCP | Python · MCP server · AI agents

Meridian runs a local API bridge that lets Pi, OpenCode, and other coding clients use Claude and Antigravity subscriptions. It includes provider management, request monitoring, a usage dashboard, and an optional macOS app.

AI Coding | TypeScript · coding clients · API bridge

OpenChatCut is an AGPL local-first AI video editor: a built-in chat agent and external MCP agents (Claude Code, Codex) edit real, still-editable multitrack timelines with transcript cuts, captions, effects, generation and Remotion/FFmpeg exports.

AI Tools | TypeScript · video-editing · mcp

BOSS Zhipin job-search CLI built for AI agents: 39 commands plus a 77-tool MCP server, emitting a stable JSON envelope. Covers welfare filtering, local shortlists, resume/JD AI helpers (Ollama/vLLM), and recruiter workflows.

MCP | Python · cli · agent-skill

ElatoAI runs realtime speech-to-speech voice agents on ESP32/Arduino hardware via secure WebSockets and Deno/Cloudflare edge functions. It supports 100+ models (OpenAI Realtime, Gemini Live, Grok, ElevenLabs, Hume, Deepgram) plus a Next.js app for creating agents, managing devices, and transcripts.

AI Agents | TypeScript · realtime-voice · speech-to-speech

MCP server giving AI agents a reverse-engineering workspace: browser automation via CDP, network interception with auto-CA MITM proxy, JS deobfuscation, WASM disassembly, Frida/Ghidra bridges, and 723 tools exposed through search-first profiles.

MCP | TypeScript · mcp-server · browser-automation

PageLM is a self-hosted AI study platform that turns documents and other learning materials into chat, notes, quizzes, flashcards, and podcasts. It supports multiple hosted LLM providers and local Ollama models, with configurable embedding and speech services.

AI Tools | TypeScript · education · study assistant

Agent skill for Codex, Claude Code, and OpenClaw that translates whole books (PDF/DOCX/EPUB) into other languages using 8 parallel subagents, with glossary term consistency, SHA-256 manifest validation, resumable runs, and DOCX/EPUB/PDF output.

Skills | Python · agent-skill · translation

NanoJev is a 0.6B Qwen3-based replica of Jev: a parallel decision model that returns probability distributions over supplied candidate actions with zero output-token decoding. Ships SFT training configs, an 18.7K-question dataset, an HTTP decision service and ViZDoom/Maze/Snake benchmarks.

AI Frameworks | Python · decision-model · parallel-decoding

Autonomous trading agent that asks an AI model (Jev via the Vercel AI SDK) for a buy/sell call every Monad block and posts a post-only limit order on Kuru's MON-USDC book. A mock momentum model enables dry runs, and an SSE server streams each block's decision, quote and fills.

AI Agents | TypeScript · trading-bot · autonomous-agent

Self-hosted Rust Server + Runner that gives cloud AI agents (ChatGPT, Claude, Codex) an MCP/HTTPS interface to your real local dev environment: files, Git, shells, tests and compilers. Code stays on your machine, with jobs, evidence, and a runtime console for human review.

MCP | Rust · mcp-server · self-hosted

Memmy is a local-first memory hub plus personal agent runtime that gives multiple AI coding agents (Claude Code, Codex, Cursor, OpenClaw, Hermes) one shared persistent memory, exposed through a desktop app, memmy CLI/TUI, a local memory service, and an OpenAI-compatible API.

AI Agents | TypeScript · long-term-memory · personal-ai

Self-hosted persistent memory service for AI agents exposing MCP plus a REST API, with a causal knowledge graph, semantic search over sqlite-vec embeddings, and autonomous memory consolidation. Drop-in for LangGraph, CrewAI, AutoGen, Claude and other MCP clients.

MCP | Python · agent-memory · long-term-memory

Zero-dependency TypeScript library that fetches A-share, HK and US stock quotes, K-lines and search results from Tencent/Sina/Eastmoney. Ships as a Node.js/browser API, CLI, an MCP server (get_stock, get_klines, ...), and a SKILL.md for agents.

MCP | TypeScript · stocks · finance

Cross-platform desktop AI assistant with chat, autonomous and multi-agent workflows, RAG, MCP connectors, skills, computer use, voice, and image/video generation. Supports major hosted model providers and local Ollama models.

AI Agents | Python · desktop assistant · multi-agent

Rust framework for building AI agents with function calling and serverless LLM tools. It provides QUIC-based tool routing, an OpenAI-compatible API, and infrastructure for geo-distributed inference.

AI Frameworks | Rust · AI agents · function calling

Skales is a cross-platform personal AI agent for desktop and mobile that can run autonomous goals, automate desktop and browser tasks, and edit code. It supports MCP, scheduled tasks, multiple AI providers, and local models through Ollama.

AI Agents | autonomous-agent · desktop-agent

A multimodal agent framework that plans tool workflows for video understanding, retrieval, editing, and generation. It supports natural-language tasks such as video Q&A, summarization, clip editing, and creative remakes.

AI Agents | Python · video agent · multimodal

Cross-platform database client built with Go/Wails + React covering ~35 SQL, NoSQL, cache, vector and message-queue data sources. Ships a built-in MCP server plus a multi-provider AI assistant so coding agents get schema-aware SQL tools while credentials stay on the local host.

MCP | TypeScript · database-client · sql-client

A desktop and self-hosted browser workspace for managing persistent terminals and parallel AI coding-agent sessions on a draggable canvas. Supports agents including Claude Code, Codex, Gemini, GitHub Copilot, and OpenCode.

AI Coding | TypeScript · AI coding agents · agent orchestration

Deprecated no-code web platform from LangChain for building, configuring, and chatting with LangGraph agents, with MCP tool connections, LangConnect RAG, and multi-agent supervision. LangSmith Agent Builder is the recommended replacement.

AI Agents | TypeScript · no-code · agent-builder

Snippai is an Electron desktop tool that uses AI (Transformers.js) to analyze screenshots: OCR text extraction, formula to LaTeX, table to Markdown, image description, code explanation, color detection, and translation.

AI Tools | TypeScript · screenshot · OCR

Self-hosted, AI-native project management platform (Jira/ClickUp alternative) where AI agents join Scrum boards as teammates, picking up tasks and co-writing BDD specs. Connects coding CLIs via ACP/MCP and extends through WASM plugins.

AI Agents | Go · multi-agent · project-management

Hosted MCP server exposing unified real-time search: general web search, vertical domain search, parallel batch queries, and full-page URL-to-Markdown extraction over Streamable HTTP. Works with Claude Code, Cursor, VS Code, Cline, Codex and more; API key optional.

MCP | mcp-server · web-search

Stripe's official AI toolkit: @stripe/ai-sdk and @stripe/token-meter for metering LLM usage into Stripe billing, a remote Stripe MCP server with OAuth, and installable agent skills/plugins for Claude Code, Codex, Cursor, and Grok.

AI Tools | TypeScript · stripe · payments

Go-native framework for production agent systems: LLM agents, LangGraph-style graph workflows, multi-agent orchestration, tools, memory, RAG, Agent Skills, evaluation, and OpenTelemetry tracing. Integrates MCP, A2A, and AG-UI for tool, agent, and frontend interop.

AI Agents | Go · golang · agent-framework

LangAlpha is a Claude Code-style agent harness for investing: a persistent per-workspace sandbox, MCP financial data tools, skills for DCF/earnings workflows, and Python code execution to process market data without bloating context. Runs deep-research or fast secretary agents across web, TUI, and chat channels.

AI Agents | Python · financial-analysis · investment-research

WFGY provides protocols and troubleshooting guides for AI reasoning, RAG, and agent workflows. Its public materials include a Problem Map for diagnosing failures, evaluation and verification resources, and a portable Goal Compiler for structuring human-AI tasks.

AI Tools | Jupyter Notebook · AI reasoning · RAG debugging

Official LM Studio TypeScript SDK for driving local LLMs: chat and text completion, embeddings, function/tool calling, autonomous agents, plus model load/unload and parameter control. Runs in Node and the browser, aimed at developers using local models instead of a cloud API.

AI Tools | TypeScript · sdk · local-llm

An MCP server for generating and editing images and videos through MeiGen Cloud, OpenAI-compatible APIs, or local ComfyUI. Includes ecommerce-focused image skills, a library of over 1,400 prompts, and parallel task orchestration.

MCP | TypeScript · MCP server · image generation

Nimbalyst is an MIT-licensed Electron desktop IDE/workspace for running parallel coding agents (Claude Code, Codex, OpenCode, Copilot) in isolated git worktrees, with visual WYSIWYG editors, diff review, kanban session tracking, task trackers, MCP client support, and an iOS companion.

AI Coding | TypeScript · coding-agent · multi-agent

MiniMax Code (mcode) is a TypeScript terminal coding agent that reads files, inspects diffs, runs shell commands and tests, with permission modes and sandboxing. It runs on a MiniMax account or BYOK OpenAI/Anthropic-compatible providers, supports MCP, ACP, subagents, plugins and a headless CLI for CI.

AI Coding | TypeScript · coding-agent · terminal

Docs MCP Server is a self-hosted MCP server that scrapes and indexes documentation from websites, GitHub, npm, PyPI and local files, then serves version-specific, embedding-backed search results to AI coding clients like Claude, Cline and Cursor.

MCP | TypeScript · documentation · rag

MEX stores shared project memory, code-linked knowledge, and teammate handoffs in a Git repository so engineers and coding agents can reuse context across sessions. It provides a CLI, local web Hub, and integrations for Claude Code and Codex.

AI Coding | TypeScript · coding-agents · project-memory

Rust tool that turns a markdown folder into a knowledge graph, exposing it two ways: an LSP server for editors (VS Code, Neovim, Zed, Helix) and an MCP server plus CLI for AI agent memory and structured retrieval. Local-first, git-friendly, no database.

MCP | Rust · ai-memory · knowledge-graph

Happier is a web, desktop, and mobile client for running and coordinating coding agents such as Claude Code, Codex, and OpenCode across your machines. It supports parallel sessions, cross-agent delegation, remote control, code review, and self-hosting with end-to-end encryption.

AI Coding | TypeScript · coding-agent-orchestration · multi-agent

A PHP framework for building generative AI applications, with support for multiple LLM providers, embeddings, vector stores, agents, and question answering. Integrates with Laravel and Symfony.

AI Frameworks | PHP · generative AI · LLM

Telegram MCP server built on Telethon that exposes 80+ tools so MCP clients like Claude and Cursor can read chats, send and edit messages, manage groups, contacts, media and folders, and transcribe voice notes.

MCP | Python · telegram · telethon

Cross-platform LLM chat app for roleplay and storytelling. Supports multiple model providers, custom characters, group chats, lorebooks, long-term memory, plugins, and chat media.

AI Tools | TypeScript · LLM chat · roleplay

Local-first long-term memory runtime for AI agents: Markdown files are the source of truth, a rebuildable SQLite/FTS5 index provides ranked path-based recall, and a sleep-time Manage layer consolidates memories. Works via CLI or MCP for Claude Code, Codex and any shell-capable host, with no API key.

AI Agents | Python · agent-memory · long-term-memory

Rust CLI and C-ABI library that gives agents reliable computer use on macOS, Linux, and Windows by reading native accessibility trees instead of pixels. Offers ref-stable snapshots, progressive token-saving traversal, session tracing, and CDP interop for Chromium apps.

AI Agents | Rust · computer-use · desktop-automation

Local-first desktop AI agent (Rust/Tauri) with persistent layered memory, autonomous Goals, dynamic Workflow/Loop execution and multi-agent orchestration. Includes MCP client, skills, RAG knowledge space, computer/browser control, and headless server mode for NAS or cloud.

AI Agents | Rust · agent · desktop-app

Official Python implementation of Apple's MobileCLIP and MobileCLIP2 vision-language models, with pretrained checkpoints, inference, training, and evaluation code. Includes lightweight variants designed for fast image-text inference on mobile devices.

AI Tools | Python · vision-language · image-text models

Free, hands-on course materials for using AI coding tools and agents across the software lifecycle. Learners build, test, deploy, monitor, and extend an application with agent workflows, MCP, reusable skills, and subagents.

AI Coding | Python · AI coding agents · AI-assisted development

iMCP is a macOS menu-bar app that runs a local MCP server exposing Messages, Contacts, Calendar, Reminders, Maps, Location, Phone, Weather, and Shortcuts to MCP clients like Claude Desktop, Claude Code, Cursor, and Amp over stdio with per-service macOS permissions.

MCP | Swift · mcp-server · macos

TypeScript CLI control plane that gives coding agents (Claude Code, Codex, Gemini CLI, Cursor, and more) one shared config, a live session console, cross-agent messaging, local SQLite memory over MCP, and composable engineering skills like dev-lifecycle and verify.

AI Coding | TypeScript · coding-agents · multi-agent

A Claude Code / Codex skill file (SKILL.md + embedded Python) giving AI coding assistants zero-auth access to US stock data: CBOE options Greeks and 0DTE flow, FINRA short volume, SEC EDGAR filings and an XBRL market screener, plus quotes, fundamentals and technicals. Each source is labeled with a compliance tier.

Skills | ai-skill · claude-code

Reference implementation of a controllable RAG agent where a deterministic LangGraph 'brain' plans, decomposes, retrieves and verifies to answer complex multi-hop questions over PDFs. Uses FAISS vector stores for chunks, chapter summaries and quotes, with Ragas evaluation and a Streamlit visualizer.

RAG | Jupyter Notebook · langgraph · langchain

Hivemind captures AI agent session traces, retrieves relevant memories, and turns repeated patterns into reusable skills shared across agents and teammates. It integrates with Claude Code, Codex, Cursor, OpenClaw, Hermes, and pi, with cloud-backed Deeplake storage.

AI Agents | TypeScript · agent memory · shared memory

Python OSINT agent that chains 20 investigation tools (email, username, breach, DNS, Shodan, VirusTotal) behind a natural-language REPL, CLI and web UI. Also ships an MCP server so Claude, GPT-4 or local Ollama models can drive the same tools. For authorized security research only.

AI Agents | Python · osint · security

A Model Context Protocol server that exposes China Railway 12306 ticket search to LLMs, supporting ticket queries, train filtering, intermediate-stop and transfer route lookups. Runs over stdio or HTTP and can be launched via npx or Docker.

MCP | JavaScript · mcp-server · 12306

Qwen's 7B text-to-image and image-editing diffusion model (32-layer single-stream DiT) with native RGBA transparency, up to 10 reference images, mask/local edits, and 2K output. Ships Diffusers, ComfyUI, vLLM-Omni and SGLang integrations plus prompt-rewriting checkpoints.

AI Frameworks | Python · text-to-image · image-editing

A TypeScript CLI proxy that lets stdio-only MCP clients (Claude Desktop, Cursor, Windsurf) talk to remote MCP servers, handling OAuth, custom headers, multi-session isolation, and tool filtering. Drop it into mcpServers config via npx until clients support remote servers natively.

MCP | TypeScript · proxy · stdio

vLLM-style inference server for Apple Silicon built on MLX, exposing OpenAI /v1/* and Anthropic /v1/messages from one process. Adds continuous batching, paged/prefix KV cache, structured output, MCP tool calling, and multimodal text, vision, audio, embeddings and rerank support.

AI Frameworks | Python · llm-inference · mlx