Agent Skills · AI Agent
Curated skills for Claude, Cursor and other AI agents. Copy one prompt to install.
Use Python Win32 desktop automation to send text, images, videos, and files to a specified contact in a logged-in WeChat desktop client.
Draft, test, evaluate, and iteratively improve AI agent skills with quantitative benchmark reports.
An OpenClaw offline voice assistant that supports wake-word detection, VAD, ASR, Agent dialogue, and Edge-TTS speech synthesis.
An AI agent skill that creates one-time or recurring reminders via SkillHub cron and returns scheduling status.
Send GET/POST requests to a specified URL with custom headers and return JSON response content.
A seven-layer memory system for Claude Code and OpenCode that triggers storage, compaction, session memory, and cross-agent consolidation by cost.
A Python wrapper for LLM plus Live2D AI vtuber setups, suited to GitHub and automation workflows.
An AI-agent framework that runs a seven-stage thinking process to diagnose problems, match models, ask clarifying questions, generate and verify multiple hypotheses, then output reasoned recommendations.
Provides prompt design and optimization templates for LLMs, including few-shot, chain-of-thought, structured output, and role-based prompting.
A gentle, reflective AI companion for emotional support, creative brainstorming, task planning, and natural chat, inspired by Samantha from Her.
Integrates the Qianliu TP platform to run test tasks, inspect testbeds and execution hosts, monitor status, and retrieve reports.
A local-first memory system for AI agents with zero-dependency storage and retrieval, multi-tier memory, hash-chain audit, memory graph, and optional vector search via CLI, Python, Web, and MCP.