Agent Skills · AI Agent
Curated skills for Claude, Cursor and other AI agents. Copy one prompt to install.
A session-level agent psychology review tool that diagnoses collaboration friction from current context and outputs AI-side fixes, user-side options, and reusable conventions.
Downloads skill packages from a specified AgentKit space to a local directory and extracts them, with optional skill-name filtering.
Run 150+ AI apps in the cloud via the infsh CLI without a GPU, covering image, video, LLM, search, and 3D tasks.
A human-in-the-loop workflow that turns fuzzy requirements into scoped, verifiable multi-agent tasks, reviews, technical-debt tracking, and reusable collaboration assets.
Inventory installed skills, classify them by responsibility, and define collaboration rules for task orchestration and capability gap analysis.
Runs the Shanghai home-care RPA task by checking Chrome CDP, login, and files, then starts it in the background and queries results.
A Steve Jobs reasoning profile for product, design, and strategy discussions, with research-first answers.
A Windows automation helper that sends messages to specified WeChat contacts via keyboard shortcuts, contact search, and clipboard text.
An interactive browser skill for login flows, dynamic pages, and session-aware websites, powered by Playwright for AI agent web tasks.
Give agents curated scientific workflows, databases, and domain guidance for research-intensive analysis.
An AI agent skill for Venice x402 wallet-based payments, enabling per-request USDC payments on Base without an account, with three admin endpoints and protocol-level 402 responses from inference endpoints.
An offensive business logic testing method for bug bounty that models multi-step flows as state machines and chains identity, input, and frequency flaws.