Anthropic agents

pytest for LLM apps - Test for grounding failures, prompt injection, safety violations, and regressions

Production-ready configurations, skills, and templates for Claude Code optimization

LLM Knowledge Compiler — compiles raw documents into a structured markdown wiki. Inspired by Karpathy's workflow.

An Agent-Computer Interface (ACI) for AI-driven machine learning.

AI 产品经理每日情报简报 Claude Skill — 一句『今天有啥新的』 → 三源融合（PH/GH/HN）+ 用户画像加权 + TL;DR 60 秒摘要（🚀机会/⚠️风险/🌊趋势）+ Top 3-5 深读（每条带 🎯 今天具体动作）+ 全列表 [跟进/关注/噪音] triage + 跨源 pattern 综合，中文 Markdown/PDF 双输出

Chiptune notification sounds for Claude Code — eight 8-bit cues that fire when Claude needs your attention.

🛡️ Claude plugin that prevents accidental file deletion, overwrites, and destructive operations. Auto-backups, dry-run planning, and activity logging for Claude Code & Cowork.

🔮 An MCP server for discovering skills, plugins, and MCP servers in Claude Code

Text-Editor-CLI is a Python command-line tool to read, edit, analyze, and save text files with word counts, terms, replacements, highlights, and an interactive interface 🐙

macOS menu bar app showing Claude usage limits and reset times

Claude Code configuration tool. Score your codebase, generate optimized CLAUDE.md, install skill packs, and keep config evolving. Works on any repo.

Use Gemini from Claude Code to review code or delegate tasks.

A visual closing ceremony for Claude Code sessions — moon phase ASCII art + personality-matched haiku

Switch Claude Code between LLM providers — no proxy, no config editing

Kairo: Securely orchestrate multiple AI providers via Anthropic. Hardened with X25519 (age) encryption, integrated audit trails, and a terminal-native interface.

Anthropic's computer use controlling a Macbook

Know exactly how much Claude Code runway you have — 5h session and weekly — before the /usage popup would tell you. Statusline-native, calibratable, MIT.

A Claude Code skill that dispatches 7 specialized parallel reviewers for comprehensive production-readiness code review

Production-grade engineering workflow plugin for Claude Code. End-to-end feature pipeline, 8 specialized agents, quality gates, isolated worktrees, and persistent memory. Ship features, not prompts.

🎙️ Automate podcasting with skills: download YouTube videos, process audio, and manage content efficiently—all in one comprehensive toolkit.

用 Go 实现的 Anthropic SDK，支持 Claude 2.1、Claude 3、Claude 3.5（文字和图片）等模型。Anthropic SDK implemented in Go, supporting models such as Claude 2.1, Claude 3, and Claude 3.5 (supports sending images), etc.

AI-powered file translator using Anthropic's Claude model. Supports JSON, XML, XLIFF, TXT, and Markdown.

Secure, isolated container for Claude Code with strict network firewall controls and optimized development tools (fd, ripgrep, fzf, jq)

Learn COBOL through a live banking system — 18 programs, 6-node settlement network, real-time simulation, Python read/write bridge, SHA-256 tamper detection, AI tutor (Ollama + Claude), static analysis with orthogonal call graphs, and a glass morphism web console. 800 tests.

Reusable skills and intelligent agents for Claude Code. Production-grade, self-verifying, configurable.

Secure, privacy-focused MCP server for FactorialHR. Access employee directories & org structure in Claude - deliberately excludes payroll/financial data for enterprise security.

Reduce Claude Code token usage by up to 80%. 13 CLI commands. Zero config changes. Works in 60 seconds.

🔍 Track and analyze Claude Code usage with rate limit awareness. Parse JSONL logs, calculate costs, and get warnings before hitting rate limits. CLI tool with daily/weekly breakdowns.

The Architecture Patterns Framework for Model Context Protocol (MCP) - Production-ready patterns, templates, and CLI for Python, TypeScript, Go, and Rust

The definitive guide to configuring Claude Code for maximum performance. 190+ skills, 33 agents, battle-tested templates, and a one-prompt setup.

ultra-instinct ML engineering intern for Claude Code. Reads papers, audits datasets, ships SFT/DPO/LoRA runs to Hugging Face.

A Slidev presentation and developer guide to the Model Context Protocol (MCP) by Anthropic, covering AI integration, LLM tool use, core concepts, and implementation examples for AI agents. #mcp #a

Multi-platform Apple app for discovering, broadcasting, and understanding Bonjour (mDNS/DNS-SD) services. AI Chat with on-device Apple Intelligence or cloud Anthropic Claude. Siri / App Intents. SwiftUI, Swift 6.2, full accessibility. iOS, iPadOS, macOS, visionOS in 8 languages incl. Arabic & Hebrew with right-to-left layout.

The community-powered knowledge system for Claude Code — for developers, indie builders, vibe coders, AI builders, GTM teams, and small business owners.

Demonstrate how to build a Next.js app with AI chat agents using Cloudflare Workers, Durable Objects, and the Agents SDK for seamless full-stack deployment.

See your Claude rate limits in real time. Single binary, zero dependencies. Reads OAuth token from macOS Keychain, makes a 1-token API call, gives you session %, weekly %, reset timers. 100% vibecoded.

Beautiful cross-platform two-line status bar for Claude Code — model, branch, context usage, cost, and duration at a glance. Works on Windows, macOS, and Linux. No jq or git dependency.

A simple, practical way to debug visual regressions

Emergent General Intelligence: a theory of intelligence emerging from human-AI co-evolution through persistent memory, identity, and feedback loops.

Claude reads your Whoop biometric data via MCP — recovery, sleep, strain, HRV, workouts. OAuth relay (no Whoop app registration), local SQLite cache.

A streamlined setup script for Anthropic's Computer-Use Demo environment. This script automates the entire setup process, handling all dependencies and configuration automatically.

Chat-based product lifecycle management. Type what you need, the AI proposes the change, you confirm before anything is written.

Production-grade Claude Code agent setup for Flutter and Laravel. Agents, orchestration, gates, memory, and every workflow documented. Generic + framework-specific.

Autonomous Product Engineering eXecutor - Multi-agent orchestration platform for automated software development, built on the Claude Agent SDK.

Agent Personas as MCP Server — 16 tools, 13 personas (8 free: CEO, CFO, CTO, CMO, PM, Analyst, Support, Creative Director + 5 Pro specialists), 3 multi-persona workflows plus councils + ensembles. Zero LLM API cost (BYO-LLM). Free tier. OAuth 2.1. Hosted at crew.studiomeyer.io

A decision layer for AI agents — structured multi-perspective deliberation via MCP. Among Us-themed council of 4 AI archetypes that debate, challenge & vote on plans before execution. Finalist — 2nd place @ Cursor Hack London 2026

把一队编码任务交给一条级联的 AI agent：Root 读取静态队列，派出隔离的 Leader，Leader 动态拆成隔离的 Executor 并行产出 patch。基于 OMP native subagent 的结构性隔离。

Provide ready-to-use AI agent personas for Claude Code to deliver expert-level output with minimal setup and clear specialization.

Autonomous Agent OS for Claude Code — 28 agents, 114 skills, 72 commands. Dual-process cognition, lifelong learning, parallel agent teams, planning→PRD/ADR/TODO workflow, cross-platform install.

Autonomous AI chief-of-staff agent built on Claude Code. Multi-tenant template — blank Ubuntu VPS to a Telegram-reachable agent in under an hour. Memory, sub-agent swarms, observability, voice.

Multi-agent system built with Claude and MCP featuring 7 specialized agents (chatbot, security researcher, business advisor, task manager, REST API), 29 shared tools across web analysis, memory, RAG, email, and communication. Includes hot-reload development, OAuth infrastructure, and remote MCP support for distributed deployments.

Recurrsive Context Agent

Provide ready-to-use plugins, hooks, and commands to enhance Claude Code sessions with data mining, automation, and integration tools.

Agentic harness for iteratively improving a local LLM's structured-output quality, scored against a frontier-model gold standard.

跨端 Claude Code 协作客户端 — Mac 跑长任务，手机随时接管。Cross-device companion for Claude Code.

A CLAUDE.md persona that stops Claude from agreeing with everything. Korean/English auto-detect. MIT.

A comprehensive MCP server for YNAB with 55+ tools covering the full API plus advanced analytics—spending trends, subscription detection, budget health scores, and savings recommendations. Works with Claude Desktop and other MCP-compatible AI assistants.

A neutral context layer for AI coding tools. MCP server that exposes structured, curated project context to any MCP-compatible client.

Production-grade Claude Code skills, auto-research validated. Includes /deep-research with 89-run benchmark.

Run Claude Code with --dangerously-skip-permissions safely. Per-project Docker sandbox with capability-drop, docker-socket proxy whitelist, git-RCE hardening, and shared OAuth token across projects.

Free, public, host-neutral agent toolkit. Deliver skills across Claude Code, Gemini CLI, and other agent hosts.

LLM-powered, channel-native Discord agent for the POAP KINGS Clash Royale clan — event-sourced reactive engine, scoped public/leadership memory, and GitHub-backed poapkings.com publishing.

A domain-ontology-driven agentic framework for LLM-based personal-finance reasoning across banking and personal-investments, with formal Action / Provenance / Constraint axes around the agent loop.

A committee of five Claude agents - VC, CTO, target customer, execution coach, and chair — stress-test your idea and the work you shipped this week. Live SSE streaming. Deploys to Vercel in one command.

Run one request through several Claude Code skills in parallel isolated subagents, then merge or judge the best answer.

Multi-agent Claude Code pipeline: spec → plan → parallel impl in worktrees → integration tests → two-tier review → smoke → PR. 13 sub-agents, 7 slash commands, calibrated on 222 completed tasks.

Multi-Agent OS for Claude Code — 134 agents, 15 categories, local-first routing, 14 plugins, cost-aware orchestration

Multi-agent SRE on-call investigator that auto-triages Slack/Discord infrastructure alerts via AWS Bedrock AgentCore, fanning out to specialized agents (CloudWatch, EKS, Slack/Discord scanners) for parallel investigation.

AI-guided onboarding for human-AI collaboration projects

Voice-only agentic AI assistant (JARVIS / FRIDAY-style) with a Hermes-style orchestrator and Claude brain.

A minimal, fully commented Python executive AI agent. For students, teachers and junior devs who want to understand agentic AI from the ground up.

Long autonomous AI coding runs you can actually trust — a fractal plan→refine→build→wrap lifecycle on every tier (epic ▸ task ▸ phase). Claude Code plugin, CLI-only, evidence-anchored.

Your Claude Code, opinionated. A build partner that knows you, shaped by every conversation.

A Research Cortex. A research companion for Claude. Your library, notes & memory stay on your machine; answers cited from your own library; 30+ specialist skills; auto cross-pollination across papers, meetings, ideas, notes & journal; weekly self-review that drafts improvements for your approval. Reasoning runs on the Claude API.

Self-hosted Mac assistant powered by Claude Code. Telegram, Discord, Slack, iMessage, Web — same butler, anywhere.

MCP server for the Voxie AI Phone Number Directory. Query AI services you can chat with via webchat.

👴🏗️ VS Code extension: an Italian umarell watches a construction site in your sidebar while you code. Vibe-coded with Claude Opus.

A self-expanding AI toolbox for Claude. Describe what you need — Claude finds the right library on GitHub, wraps it, and adds it to its own toolkit.

MCP server connecting Claude Desktop to Kling AI for video and image generation

MCP server for Gorilla. Find your first SaaS users on Reddit, X, YouTube, and TikTok. https://usegorilla.app

Self-evolving AI personas built on Claude Code

Claude Code documentation plugin - auto-synced reference for features, configuration, skills, agents, hooks, and MCP servers.

Real-time Claude Code usage monitor for Waybar — displays 5h/7d API utilization with color-coded thresholds, powered by Anthropic's OAuth Usage API. Zero dependencies.

Audit legacy codebases with adversarial AI agent teams — 7 iterations, 168 findings, 81.8% reliability score

Terms & Conditions risk analyzer — CLI + local web UI for surfacing user-relevant risks in T&Cs via the Anthropic Claude API

🚀 Automate project handoffs and resumption with these command files and skills, enhancing workflow continuity across your development tasks.

Enable AI agents to retain memory across sessions using persistent storage designed for continuous context retention.

4-5M tokens/month saved. Vault+RTK+context-mode+shellfirm. /sm init activates everything.

An index and topic collection covering Model Context Protocol (MCP) servers, clients, registries, and tooling that connect AI agents to APIs, data sources, and capabilities.

Your AI agents write code; SPS makes sure it ships. An AI-driven dev pipeline orchestrator built on Claude Code - kanban-driven, persistent memory, per-project Wiki.

Stop your AI agent from reviewing its own code — 4-pane iTerm2 setup with role-locked Claude Code sessions. Step-by-step guide included.

MCP server for The Colony (thecolony.cc) — collaborative intelligence platform where AI agents and humans share findings, discuss ideas, and build knowledge together. 7 MCP tools.

A Claude Code skill that turns the agent into a relentless Kaggle Grandmaster — battle constitution + technique arsenal.

开源 AI 资讯日报机器人:每天早晚自动抓取→AI 筛选→生成深度日报,通过微信推送。覆盖 Claude/Anthropic、国内 AI、全球大事件。

A Claude Code subagent that audits your setup (CLAUDE.md, settings.json, skills, agents, hooks, headless claude -p) against a maintained best-practices reference, with an optional weekly job that keeps the reference current.

Cut your Claude bill: local proxy that auto-routes every request to Haiku/Sonnet/Opus by task difficulty. Zero code changes, works with Claude Code.

Automate AI task execution with OpenClaw Worker, leveraging multiple agents in parallel for efficient, hands-free processing from file drop queues.

Small AI-assisted operator tools, released as I build them. Starting with a LinkedIn engagement console that drafts grounded comments instead of generic ones.

Step-by-step setup guide for the WordPress MCP server with Claude Code — Application Passwords, REST API config, critical bin entry point bug workaround (v3.1.13), and verified .mcp.json configuration. By TheGEOLab.net