Explore skills

Structural AI for Macroeconomic Intelligence

PRE Engineering sets up a Plan-Review-Execute multi-agent collaborative framework for any project. Three AI agent roles (Planner, Executor, Reviewer) work together through a shared collaboration log to continuously deliver project goals.

Expert skill for the German research allowance (FZulG). It guides applicants step-by-step through the entire process, from applying for the BSFZ certificate and formulating project descriptions to checking applications for plausibility, avoiding common errors, and maximizing the success rate.

This skill uses computer-use to control desktop WeChat, allowing users to send and reply to messages in group or private chats. It activates for explicit WeChat requests or implicit chat commands, generating natural messages based on chat history and relationship.

This skill transforms any content into a Socratic dialogue, featuring a believer's presentation, Socrates' deep questioning, and a final reversal. It activates only when users explicitly request a 'dialogue/Socratic style', not for general writing or polishing.

This skill performs single-CVE applicability triage, evaluating if a CVE applies to the user's environment by comparing it against a maintained environment profile. It activates when a user inquires about their CVE exposure or wants to assess a CVE against their environment.

Generate, fix, explain, and review DQL. Any final executable DQL must first pass dqlcheck validation item by item before being delivered.

A Gaokao essay writing coach with 8 modes: topic analysis, idea refinement, outlining, sample essays, grading, full book grading, corpus search, and timed practice. It employs core methods like the 'Sanli' topic analysis and 'Five sentences one paragraph' techniques.

A Claude Agent Skill providing comprehensive Rails best practices from rails-bestpractices.com.

Diagnoses organizational health against Brian Chesky's framework, detecting the division → politics → bureaucracy → complacency arc that kills founder-led companies. Use when a founder suspects their org structure is slowing them down.

Configures and activates the conda/mamba/pixi environment for Python commands. Checks the project's CLAUDE.md for environment configuration.

A two-tier project memory system maintains a global Claude memory snapshot and a per-project STATUS.md file, capturing session state, actions, and blockers to ensure context survives long absences. Its WRITE mode is triggered by user commands like "save this" or "checkpoint [project]".