Agent Skills
Curated skills for Claude, Cursor and other AI agents. Copy one prompt to install.
A pure Python, security-first memory layer for AI agents with injection detection, PII masking, hybrid retrieval, and lifecycle management.
Under scoped data sources, it searches hot works on WeChat Official Accounts, Douyin, and Xiaohongshu, expands keywords, validates samples, and outputs reusable topic and angle leads.
Generate structured lesson plans by subject, grade, and class duration, including lesson flow and Word export.
Configure holdings and falsifiable thresholds, then scan announcements, reports, operations, and industry events to generate a structured HTML weekly report.
Refines .docx novels for Tomato review by fixing policy issues, removing AI tone, and checking 80k-word evaluation and pacing.
Guides DingTalk office workflows for work notices, group bots, approvals, and attendance with credential checks, token caching, and error troubleshooting.
An interview practice simulator that generates role-specific behavioral, technical, and case interview questions with follow-ups, feedback, and improvement suggestions.
Interrogate design decisions one question at a time while aligning glossary, ADRs, and code facts with stdlib validators.
Generates a seven-dimension stock research set and an independent analysis with clear short-term and long-term conclusions.
Parses clinical notes and lab data to generate structured suspected diagnosis, differential diagnosis, misdiagnosis risk, and missing-information reports.
A financial research and quant toolkit that integrates A-share data, multi-market data, and report workflows for retrieval, analysis, and output.
Applies Inamori's amoeba management method to food service operations, turning problem framing, analysis, solution design, and review into data-driven insights.