Skill UI

Explains any startup equity, legal, or term sheet question—SAFE notes, convertible rounds, and preferred stock—while covering both investor and founder lenses and flagging red flags or negotiable points.

This skill generates user-friendly, public-facing release notes (changelogs) for App Stores. It analyzes the raw Git history since the last tag, intelligently filtering out internal commits (like refactors, CI updates, or build scripts) and grouping visible changes into concise, benefit-focused bullet points suitable for a "What's New" section.

This tool systematically structures and derives complex research formulas and theoretical models. It takes scattered equations, messy notes, and ambiguous ideas, transforming them into a coherent, paper-ready derivation package. Use it when you need to build a rigorous theory line, organize explicit assumptions, or clarify the conceptual object underlying a set of formulas.

This skill is designed for running comprehensive evaluations of Hugging Face Hub models on local hardware. It supports multiple evaluation frameworks, including `inspect-ai` and `lighteval`, and provides flexibility in backend selection (e.g., `vllm`, Transformers, or `accelerate`) to optimize for performance and compatibility. Use this for local GPU benchmarking, smoke testing, and backend selection, but note that it does not handle remote job orchestration or result publishing.

NOTE: This skill is DEPRECATED and should no longer be used for creating new posters. The functionality has been superseded by /paper-poster-html. Users must immediately migrate to the new HTML/CSS pipeline for accurate, measurement-gated poster generation, avoiding the legacy LaTeX flow.

Manim is a powerful Python library used for creating high-quality mathematical and scientific animations. This skill facilitates the generation of educational videos and complex visualizations for presentations. It is ideal for researchers, educators, or content creators who need to visualize complex concepts in an engaging animated format. Note: This specific file is related to localization/translation.

Permanently deletes the context-mode knowledge base and resets session statistics. Users can choose between a project-wide wipe (deleting all indexed content and session data) or a targeted session wipe, ideal for isolating polluted data without losing project history. Note that this action is highly destructive and irreversible.

This tool compresses natural language content (notes, todos, etc.) into a highly condensed "caveman" format. It dramatically reduces the token count by removing filler words and redundant phrases, while meticulously preserving all technical details, code blocks, URLs, and original document structure. Ideal for optimizing large knowledge bases or long-form notes before AI ingestion to save input costs.

CLI helper that forwards a Gmail message to new recipients while preserving the original thread, attachments, and formatting; supports notes, HTML content, CC/BCC, attachment overrides, dry-run inspections, and draft saving for streamlined workflow automation.

Command-line wrapper for Google Keep allowing creation, listing, retrieval, deletion, and attachment download of notes through `gws keep`.

Automates converting a Gmail message into a Google Tasks entry via the gws workflow command; ideal for turning important emails into actionable tasks by mapping the subject and snippet to task title and notes.

A utility for interacting with the Lark/Feishu communication directory. It allows users to search for employees by name or email to obtain an open_id, or use an open_id to retrieve comprehensive profile data, including department, contact details, status, and signature. Ideal for preparing information before sending messages or scheduling events. Note that department tree traversal and organizational structure mapping require native OpenAPI calls.