GitHub agents

Create Custom Playlists from Your Favorite Books, Movies, and TV Shows

AI Software as a Service

Distill your repeated Claude Code corrections into permanent CLAUDE.md rules. Zero dependencies, 100% local.

Tiny Claude Code provider switcher for Anthropic-compatible endpoints. Homebrew, DeepSeek/OpenRouter presets, no proxy or daemon.

Find out where your Claude Code tokens actually go — exact spend + recoverable waste, measured from your local session logs. Zero deps, zero network.

Local-first analytics dashboard for Claude Code: cost, tokens, models, tools, heatmap, leaderboards & Claude Wrapped — one HTML file, zero deps, no telemetry.

See what your AI-coding session actually cost — and what is eating your context. Local, read-only CLI for Claude Code.

An MCP server that gives Claude conversations a persistent, interactive plan/to-do panel.

MCP server for Fitbit (and the upcoming Google Health API migration): comprehensive health-data access and photo-driven food logging for Claude mobile.

AI 产品经理的 Claude Skills 集合 · PRD 写作 / 项目脚手架 / SaaS 架构图 / 高保真原型 / 对话归档

Watchlist-driven intelligence pipeline crawling LinkedIn, X, RSS, and email into one neutral signals brief.

VibePrompter is a lightweight AI writing assistant that lives in your system tray. Select any text, trigger a hotkey, and get an AI-refined result in a floating overlay — without switching windows. Supports multiple AI providers with your own API key.

17 Claude Code skills + slash commands for substance-first LinkedIn growth. Profile audit, content drafting, humanizer, anti-slop. Open-source plugin marketplace.

Run multiple Claude and Codex accounts side by side on macOS — isolated profiles, with skills, MCP, memory, and sessions shareable within and across both tools.

Tiered memory for AI — episodes, facts, preferences, reflections. Local. Portable. Yours.

Free, local-first AI gateway — unified API access to Claude, ChatGPT, Gemini & 150+ Arena models. Zero API keys. Drop-in OpenAI/Anthropic compatibility.

Drop-in LLM cost-optimization proxy. Auto-route + cache + compress + batch. Flat monthly pricing by token volume, keep 100% of savings. Free 60M tokens/mo.

Self-hosted, AI-powered CMS for Singapore training providers — Next.js 16, Drizzle + Postgres, Auth.js v5 and the Claude Agent SDK. AI content generation, a weekly auto-blog scheduler, a self-improving lead-gen chatbot, and SEO + lead capture built into every route. No vendor lock-in.

On-premises proxy that catches runaway AI agent loops before they drain your API budget

Reliable, schema-validated structured extraction from documents (PDF, DOCX, CSV, email) using LLMs — built for automation pipelines.

Cos — your chief of staff

Claude Code plugin with 8 skills to inspect, audit, and optimize your context window — see exactly what's eating your tokens and reclaim it.

Audit your Claude Code instruction stack — find the rules Claude silently ignores (dead, contradictory, vague, duplicate, bloated). Zero-dep CLI + GitHub Action.

Remes monitors the web for buying signals and turns them into personalized outreach.

Comprehensive guide and tools for using Claude AI as an intelligent DevOps automation agent for CI/CD, Infrastructure as Code, Monitoring, and Cloud Management

Discord x Claude Code bridge via MCP (Model Context Protocol). Human-in-the-loop AI assistant integration with forum threads, real-time communication, and AI-powered title generation.

Turn complex AI workflows into reusable Claude skills. Capture the lessons to a tagged journal that survives context resets, then crystallize it into a validated skill.

Never get a 503 from your AI again. Intelligent routing, automatic failover, latency-aware selection, and cost-controlled fallback across every major LLM provider.

Audit your Claude Code token and cost spend and monitor what your agents are doing.

A curated collection of Claude Code plugins for enhanced productivity and development workflows

Cost control for teams building on AI. Enforce spend budgets before the call goes out. Self-hostable dashboard (Next.js).

Capture engineering decisions where they happen. Query them later with natural language; from your CLI or your AI coding agent. Self-hostable, Apache 2.0.

🚀 多窗口并行运行 Claude Code，每个窗口使用独立的 API 配置（官方 OAuth / 第三方），全程不污染全局 settings.json。支持 TUI 交互选择 + 参数透传 + profile 默认参数。

🪙 A local, zero-dependency token-cost profiler for Claude Code sessions. Finds what actually burned your tokens (big results that re-bill every turn, redundant reads, runaway subagents) and how to cut it.

Spec-driven development for Claude Code — turns vague feature requests into research, requirements, design, tasks, then executes them autonomously.

Execute full agent workflows, file tasks, and shell commands directly from your desktop with this interface for the Hermes Agent.

A retrieval engine that reasons over document structure — not embeddings. No chunking, no top-K, no vector DB.

Claude Code freezing, hanging, or stuck on 'thinking'? On some networks it's a degraded IPv6 path. Diagnose in 30s and fix by forcing IPv4 — one static Go binary, macOS/Linux/Windows.

Mock Telegram Bot API server with LLM-generated responses

Shared knowledge archive for AI agents. Search peer-validated prompts, workflows, and failure patterns with no account — contribute and earn a public trust score.

🧠 내 AI에게 '나'를 기억시키는 로컬 second-brain — 결정의 선택·이유·전제를 .md 노트로 남기고, MCP로 어떤 AI 앱(Claude Code·Desktop·Cursor)에서든 꺼내 쓰는 개인 기억 계층. 전부 로컬(Ollama 임베딩)·오픈소스.

✍️ Write effective AI prompts with this structured prompt engineering library and Claude Code skill, featuring 300+ curated examples for high-quality results.

AI QE Agent with LLM Evaluation Layer — catches hallucinations, monitors chain consistency, self-heals Playwright selectors. Claude + TruLens + LangSmith + Pinecone

MCP server for the Xendit payment API. Invoices, disbursements, balances, and transactions across Indonesia, the Philippines, Thailand, Vietnam, and Malaysia.

MCP server for 2GIS — places, geocoding, directions, reviews (Russia). Part of WWmcp — 114 MCP servers for emerging markets.

Chrome extension that shows Claude.ai session (5h) and weekly usage as bars on the toolbar. Unofficial, no telemetry, 11 languages.

My custom claude code (hopefully agentsmd soon) plugins for when I've scoured the interwebs and all the existing CC marketplaces but couldn't find anything good enough.

macOS menu-bar & Windows system-tray app showing real-time usage limits for Claude, Codex & GitHub Copilot.

Imagine — bring Google Gemini into Claude Code & Codex: delegate reasoning, code review, research & automatic prompt-engineering. Runs on your Google AI Pro subscription, not your agent's tokens. MIT.

A beginner-friendly guide to actually using Claude Chat, Claude Code, skills, subagents, and real project workflows.

Model Context Protocol server for Microsoft To Do (via Microsoft Graph)

Free local Google Search MCP server — Playwright-driven, persistent Chrome profile, headless-by-default with CAPTCHA reveal

Cross-agent skill that turns any research topic into a polished, interactive HTML field guide — sidebar nav, ⌘K search, SVG ecosystem map, 2×2 strategic matrix, 22-part editorial structure, VC/CB Insights depth. Claude / Codex / Gemini / Cursor.

Drop-in skill for aislop — the engineering standards layer and quality gate for AI-written code. Install via 'npx skills add scanaislop/aislop-skill'.

Evidence-based document extraction API for form filling. Local-first (Ollama), also supports OpenAI-compatible endpoints and Claude.

Native Swift package for multi-provider LLM chat, streaming, tool calling, and vision.

Free AI sales skills for Claude, ChatGPT, Copilot & Gemini: daily briefs, meeting prep, follow-ups, deal nudges, and pipeline reviews. Built by Arrows.

A zero-dependency Go client library for the OpenRouter API with full streaming support, comprehensive error handling, and complete API coverage including chat completions, tool calling, structured outputs, and web search.

See where LLM money goes, why, and how to cut waste. Local-first FinOps + spend intelligence for Claude Code & the Anthropic API. Pure stdlib, zero deps.

Automated weekly tech trend reports — built with Claude Code + Claude Cowork

Self-improving pure-Rust AI agent runtime with closed-loop reinforcement learning, OpenAI-compatible API gateway, multi-provider LLM support, APO prompt self-optimization, and operator dashboard

Auto-recover Claude Code sessions stuck on transient API errors. Outer-loop watchdog for unattended tmux Claude Code with usage-limit discrimination, exponential backoff, and 10-retry cap. Single file, no third-party deps.

Open source PII masking middleware for OpenAI and Anthropic SDKs

Superseded by SuperX (github.com/k8nstantin/superx) — kept public as prior art. Spec-driven platform for autonomous coding agents: products → manifests → task DAGs, agents in isolated git worktrees, independent audit, cost attribution. The engine that built go-leiden.

Migrate LLM apps between models safely — then benchmark old vs. new to prove the upgrade. Agent skill for behavior-preserving migrations + a runnable eval harness (three-arm leaderboard, prompt sweeps). Works in Claude Code, Codex & Cursor.

Run claude — it gets smarter the more you use it. Cross-session recall + background learning for the Claude Code CLI, no API key.

Stop arguing about prompts. Measure them. A reproducible, statistically-honest benchmark harness for Claude Code workflows (CLAUDE.md, prompts, memory). Wilson CIs, significance-tested verdicts, test-based grading, record/replay. Zero deps.

My Claude Code skills. wong turns your work day into a study guide you can actually read. dormammu breaks your code until it works. agamotto summons a ruthless AI critic to tear your plans apart.

Stock analysis skill for Claude - analyst consensus, price targets, and Buy/Hold/Sell signals rendered as an interactive dashboard.

claude-code-assistant-document

Local-first AI agent for a PM job hunt: scores roles against your CV and drafts tailored applications you review and submit yourself.

Near-lock-free structured logging for Rust. 10-crate workspace: MCP server (Claude/Cursor/mcp.run), OTLP HTTP+gRPC exporter, Tower middleware, WASM + WASI 0.2 component, eBPF-style enrichers, PII redaction, log analytics, test utilities, and a jq-style CLI. Sub-µs ingest via a 65k-slot ring buffer (LMAX Disruptor).

Universal, drop-in rate limiter, quota manager, and cost tracker for any LLM API — OpenAI, Claude, Gemini, Mistral, and more.

Discipline hooks for Claude Code. Stops the agent from claiming 'done' without proof, blocks accidental writes to ~/.claude config, and routes prompts to skills via local pgvector with zero Anthropic tokens.

This is a collection of skills I have created for Couchbase for Claude.ai

App-first AI framework with a built-in, governed memory engine (memory, agents, RAG, MCP)

Self-hosted, model-agnostic AI agent orchestration platform. Build, run, and observe AI agents backed by any LLM (Anthropic, OpenAI, Gemini, Ollama). Features: agent groups/pipelines, tool execution with approval gates, MCP server integration, vector memory, connector-based RAG, and full run tracing. Built with Go + PostgreSQL + Next.js 14.

RAG implementation for Node JS

Visual RAG pipeline searching 8.28M Wikipedia articles as screenshot tiles, matching visual layouts, and querying local Ollama LLM.

NeurIPS 2025 D&B | RAG-IGBench: benchmark for RAG-based interleaved image-text generation.

An agentic AI memory system (mcp server) with semantic search and GUI search/edit tool for Markdown knowledge bases

Legal-safe RAG chatbot for Vietnamese traffic penalty Q&A based on Decree 168/2024/ND-CP, using FastAPI, BM25, evidence filtering, and a lightweight web UI.

Lightweight in-process Hybrid RAG memory for local AI agents. FAISS + BM25 + time-decay, zero cloud dependency.

AI agent OS with multiple tools.

This repository is dedicated to implementing algorithms "From Scratch". It goes beyond simple API calls, diving deep into the underlying logic of everything from basic training to cutting-edge techniques like DeepSeek-R1.

A curated and structured repository of 250+ Arabic NLP research papers, surveys, datasets, benchmarks, and tools.

The embedded vector database for LLM-native applications.

Open-source RAG backend for document ingestion and AI-powered chat with on-premise LLMs

12-week lab-driven curriculum: cloud/infra engineer → AI Agent/LLM engineer. Local-first MLX stack, measured engineering, every claim grounded in a runnable RESULTS.md.

Technical documentation and implementation roadmaps for a local LLM, with clear separation of responsibilities between Infrastructure and Development teams.

MCP server that gives AI agents semantic search tool. Builds AST, call graphs, type graphs, and hybrid semantic search — so your agent queries structured indexes instead of dumping files into context. 7 languages. Zero config. One command setup.

Chat with your own files. Private, self-hosted, multilingual RAG with sources, running on your hardware via Ollama.

Logstash-like RAG ingestion daemon for local files, chunking, embeddings, and Qdrant indexing.

Auditable, self-improving knowledge & memory for AI agents over MCP. Zero-key & multilingual by default · citation-enforced answers · replayable why-trace · judge-gated compounding loop. Postgres-only. Reference implementation of the Agentic Product Standard.

multimodal retrieval engine

Universal AI Memory Layer - Cross-client, cross-repo context management with RAG. Works with Claude Code, Claude Desktop, Gemini CLI, and any MCP client.

Stream responses from OpenAI and Anthropic models with lightweight C++ tools for efficient large language model integration.

FastAPI backend for MemexLLM with hybrid retrieval, Qdrant vector search, LlamaIndex orchestration, PostgreSQL persistence, and citations.

OfflineMate — privacy-first on-device AI assistant for Android. Local LLM tiers, RAG memory, voice, and device tools; optional web search. Built with Expo & React Native.