3. Data, Science & AI

Provides video and audio transcription using local Whisper (free) and OpenAI API (fast) modes.

Extracts frames at intervals from dashcam videos to generate compact visual summaries for motion trajectory analysis and storage optimization.

Provides GPU-accelerated human detection, tracking, and classification in dashcam footage with head covering analysis for investigative purposes.

Generates comprehensive database designs including ER models, SQL scripts, and sharding strategies from high-level requirements.

Provides AI-powered capabilities through integration with OpenAI's models for natural language processing and generation tasks.

Optimizes AI programming workflows by identifying autonomous Subagent tasks and generating tailored prompt templates to maximize coding efficiency.

Analyzes Claude Code session logs to track token usage, context window behavior, and decision patterns for debugging and optimization.

Provides RAG-based question answering from indexed vault content, including setup, configuration, and troubleshooting for RAGtriever.

Provides a unified four-tier memory system for AI agents, including semantic, episodic, and causal memory layers, enabling memory-first architecture per ADR-007.

Provides LangChain integration for constructing and managing AI application workflows using large language models.

Enables initialization of LangChain skills, facilitating setup for language model application components and integrations.

Syncs AI editor prompt libraries (agents, rules, commands) from the skill's directory to GitHub Copilot and OpenCode global configurations for seamless updates.

Builds production-ready AI agents using the Claude Agent SDK for Python, supporting workflows, custom tools, and context management.

Optimizes GPU kernels for memory-efficient large language model training using the CubeCL framework.

Provides command-line utilities for interacting with LangChain, enabling developers to build and manage language model applications.

Handles saving, recalling, and organizing user interactions and findings within an AI agent's memory system.

Generic placeholder skill imported from Anthropic, intended for Claude AI environment but lacking defined functionality.

Manages Claude Code subagents with specialized expertise, handling creation, modification, and configuration via YAML, system prompts, and tool restrictions.

Guide for managing interactions and workflows between multiple AI models to enable complex collaborative tasks.

Performs schema validation and read-only analysis in PostgreSQL to generate aggregated reports for decision-making.

Enables LangChain framework integration for constructing language model applications, including prompt chains and data retrieval systems.

Generates AI images from text prompts using Google Gemini via browser automation, featuring persistent authentication and reference image style matching.

Analyzes and develops innovations in quantum, cryptography, and algorithmic systems through research, modeling, and gap analysis to challenge intractability assumptions.

Provides Vercel-integrated access to Anthropic's Claude AI model for text generation and AI-powered interactions within projects.