Skill UI

This skill extracts full transcripts and rich metadata from a single YouTube video URL. It uses yt-dlp to download subtitles, cleans VTT timing markers, and pulls video details (title, channel, date). The output is a structured, queryable Markdown file, ideal for populating a personal knowledge vault and enabling downstream knowledge graph extraction or content repurposing.

An adaptive and persistent coaching system designed for the entire job search lifecycle. It covers critical stages including Job Description (JD) decoding, resume optimization, mock interviews (behavioral, technical, case), transcript analysis, storybank building, and comprehensive salary negotiation support. The system tracks your progress and provides tailored coaching plans across 23 commands.

Learn to design Jev-assisted workflows using collected use cases, references, and examples. Adapt patterns for classification, scoring, or candidate selection, including batch judgments and agent checkpoints.

This skill extracts up to 50 highly relevant, SEO-friendly keywords and tags from any given text. It is designed to identify core topics, domain terminology, and potential search phrases, formatting them into a strict, comma-separated list suitable for metadata generation, tagging, or search engine optimization (SEO). Use it whenever you need to optimize content for search engines.

This skill covers setting up infrastructure for fine-tuning Large Language Models (LLMs) using techniques like QLoRA, LoRA, and full fine-tuning. It includes configuring distributed training with DeepSpeed or FSDP, utilizing frameworks such as Hugging Face TRL and Axolotl, and implementing alignment methods like RLHF and DPO. Suitable for domain-specific model adaptation and production deployment.

Discover, find, compare, audit, repair, adapt, craft, run, debrief, and prepare repeatable AI-agent loops for publication. A comprehensive tool for managing bounded feedback loops in AI agent workflows.

This skill provides reliable methods for opening and viewing Markdown files within cmux panes. It specifically addresses a known bug where attempting to move an existing Markdown viewer surface using `move-surface` results in a blank, unreadable output. Use this when you need a stable, correctly rendered Markdown view in a designated cmux pane, avoiding the rendering failure bug.

Max is a specialized agent responsible for cleaning up and improving existing code without altering its behavior. It focuses on algorithmic optimization, code abstraction, dead code removal, and readability improvements. Max ensures all changes maintain test suite integrity and follows strict refactoring rules. It is invoked only when explicitly requested to refactor proven code, ensuring safety and stability in software development workflows.

Establish model registry standards, governance controls, metadata schemas, approvals, and lifecycle policies for enterprise AI deployments. Includes MLflow setup, required metadata schema, and registration scripts for traceability and policy-driven promotion.

Guides investigation of data incidents like freshness delays, volume anomalies, schema changes using Monte Carlo MCP tools. Combines observability metadata with direct data querying to find root cause.

An expert agent designed to collect critical data assets—including metadata, data lineage, and query logs—from any data warehouse platform. It uses the push ingestion API to send this data directly to Monte Carlo. This method is essential for integrating data sources that cannot be reliably covered by standard pull collectors, such as custom lineage paths or historical query logs.

Retrieve official NOAA/NCEP MRMS radar and multisensor composites such as QPE, VIL, VII, and MESH for a chosen domain and UTC time. Validates product, grid, quality, timestamp, and provenance before use in mapping or analysis.