3. Data, Science & AI

Generates and animates mathematical equations using LaTeX-based rendering with customizable color mapping and equation classes.

Analyzes project structure and dependencies to generate AGENTS.md files, enabling standardized AI agent documentation and consumption conventions.

Enables natural, culturally-aware Cantonese conversations with responsive dialogue generation and linguistic understanding.

Guides creation of Model Context Protocol (MCP) servers using FastMCP in Python for LLM integration with external APIs and services.

Automates specialist agent selection and task routing based on user requests without explicit mentions.

Generates structured custom agents for Claude Code based on user descriptions, enabling task-specific automation and sub-agent creation.

Safely evolves data schemas and type definitions during model changes, ensuring consistency in data structures and field modifications.

Debugs data storage and persistence issues, including lost data, storage problems, and state persistence bugs in applications.

Extracts specific JSON fields efficiently with jq, saving 80-95% context.

Optimizes MySQL database performance through query tuning, schema design, and configuration best practices for efficient data management.

Creates and manages ServiceNow reports, dashboards, and data visualizations for business analytics and data export.

Resumes RALPH execution from the last checkpoint, ensuring continuity in data or AI workflows.

Extracts structured data from websites for data processing and analysis tasks.

Coordinates multiple AI agents to execute complex tasks in parallel, optimizing distributed problem-solving workflows.

Enables AI systems to maintain contextual memory and learning across user sessions, improving conversational continuity and personalized interactions.

Performs semantic search on vector embeddings to find similar concepts and related entities using natural language queries.

Manages Model Context Protocol (MCP) integrations for AI agents, including external services like crawlers and sandboxes for seamless LLM workflows.

Extracts and structures documentation content from websites for building offline knowledge bases and RAG pipelines, supporting recursive crawling and content filtering.

Enables natural language querying of biomedical data catalogs to find, compare, and list sources for TCM, genomics, and compound databases.

Automates extraction of documentation content from websites to build structured knowledge bases for RAG pipelines and offline reference.

Queries PostHog for product analytics, generates insights and dashboards, and tracks AI referral traffic patterns.

Optimizes LLM interactions via Qdrant-powered semantic caching, long-term memory, and hybrid search to reduce token usage and context size.

Converts Product Requirement Documents (PRDs) into structured JSON format for processing by the Ralph autonomous agent.

Runs an autonomous development loop using a multi-agent AI pipeline to automate software development tasks.