Agent Skills · AI Agent
Curated skills for Claude, Cursor and other AI agents. Copy one prompt to install.
Drive a headless browser through CLI with natural language to navigate, click, fill forms, extract data, and return structured results with screenshots.
Sets up workspace memory, append-only work logs, and optional daily/weekly/monthly reports.
Creates, tests, evaluates, and iteratively improves Agent Skills with prompt-based test cases, benchmark runs, and triggering-description optimization.
Complete COROS MCP login, regional token caching, tool discovery, and tool calls through the COROS gateway.
An engineering workflow for patent discovery, innovation-point mining, data-backed visualization, disclosure drafting, and project archiving.
Facial emotion classification for images or videos with intensity scoring, anomaly marking, and structured reports for interaction and wellness monitoring.
A meta-level multi-agent collaboration layer for OpenClaw that dynamically creates ephemeral subagents, decomposes complex tasks, and applies review, templates, caching, and model selection.
Converts vague requirements into structured, reusable AI prompts to improve output quality and controllability.
A passive cross-session memory bridge that identifies projects, loads relevant context, records key decisions, and creates local project memory files automatically.
A persistent memory management tool for AI agents, supporting init, push, recovery, local and cloud search, file watching, and distillation hook syncing.
Let AI agents log errors, corrections, feature requests, and best practices to Markdown, then promote reusable learnings to project memory.
A sales-demo automation assistant supporting workflow orchestration, data sync, RPA, monitoring, alerts, reporting, and rollout recommendations.