Browse the library
Filters
1,146 results
oh-my-agent
first-fluke
A skills pack or full harness that verifies agent work with stop-hook gates, artifact checks and independent judges across Claude Code, Cursor and Codex.
wenyan-mcp
caol64
This MCP server allows AI agents to format and publish Markdown articles directly to WeChat official accounts, bypassing manual editing.
awesome-muse-connectors
Anil-matcha
A community catalog of 150 connector skills and Meta Muse integrations, with notes on capabilities, authentication and permissions.
Aegis
GanyuanRan
A method pack for AI agents that enforces architecture-aware planning, evidence-based verification, and technical debt tracking.
gpt-image2-ppt-skills
JuneYaooo
This Claude Code skill generates high-quality, aesthetically pleasing PowerPoint presentations using GPT-image-2, either from scratch or by mimicking existing templates.
repobrain
study8677
This plugin gives AI agents "brain" access to your code repository, enabling grounded Q&A about the codebase with file paths and line numbers.
tutor-skills
bevibing
This Claude Code skill transforms PDFs, documents, and codebases into Obsidian study vaults and generates interactive quizzes based on the content.
attention-span
alexgreensh
Provides ADHD-friendly output styles for Claude Code and similar AI agents to make their responses concise, easy to skim, and human-readable.
monitor
BetterDB-inc
Provides real-time monitoring, slowlog analysis, and audit trails for Valkey and Redis databases via an MCP server.
MCPJungle
mcpjungle
Provides a self-hosted MCP gateway to manage and consolidate multiple MCP servers, offering a single endpoint for AI clients and agents.
ai-setup
caliber-ai-org
Syncs AI agent configurations, skills, and context files across Claude Code, Cursor, and Codex, with continuous context refresh and compaction features.
jupyter-mcp-server
datalayer
Connects AI agents to Jupyter Notebooks via an MCP server, enabling agents to manage and execute code in notebooks, locally or in the cloud.
autoprompt-skill
Spielewoy
A coding-agent skill that improves reliability in agentic coding tasks by reviewing, fixing, and rechecking agent work to cut failures.
meta-ads-mcp
pipeboard-co
Allows AI agents to manage Meta Ads (Facebook/Instagram) campaigns: launch, upload creatives, update budgets, and analyze performance conversationally.
iot-dc3
pnoker
Connects AI agents to industrial IoT devices via an MCP gateway, enabling agents to act on physical devices, collect data, and manage them safely.
mcp-searxng
ihor-sokoliuk
Provides AI assistants with privacy-respecting web search capabilities by integrating with a SearXNG instance via an MCP server.
AI-Search-Hub
minsight-ai-info
Aggregates native AI search from platforms like Gemini and Grok, allowing agents to perform deep web searches and extract data from diverse sources.
ios-simulator-skill
conorluddy
This Claude skill allows AI agents to build, test, and interact with iOS applications using Xcode and the iOS Simulator, parsing results effectively.
open-reverselab
LING71671
An open-source AI reverse-engineering agent platform and MCP server for automated binary analysis, CTF, and malware research.
shanyin-screenwriting-master
Shanyin-ai
A Claude AI skill for full-format film and TV series screenwriting, covering character design, outline, scene breakdown, and complete script creation.
reverify
2akouwu
This MCP server and CLI tool prevents AI agents from hallucinating by verifying every claim against ground truth with evidence, preserving context across sessions.
skills
michaelshimeles
Agent skills and an AGENTS.md template for worktree isolation, service-layer structure, evidence-based testing and before/after screenshots in PRs.
pi-subagents
tintinweb
A pi extension that brings Claude Code-style autonomous sub-agents and workflow orchestration, supporting parallel execution and mid-run steering.
deja-vu
vshulcz
Deja-vu provides persistent memory for Claude Code, Cursor, and other AI coding agents by indexing local session history, enabling instant recall of past solutions without explicit prompting.