3. Data, Science & AI

Enables direct consultation with the Claude AI model for questions, feedback, and approach validation, supporting code reviews and decision-making.

Enables semantic search and knowledge retrieval via vector embeddings for RAG systems and agent context building.

Provides marketing analytics capabilities for tracking, attribution, reporting, dashboards, and A/B testing.

Reference for protein binder design using AF2, PyRosetta, and RFdiffusion for interface optimization and design filtering.

Analyzes and validates GLTF/GLB file structures for 3D robot models, identifying mesh hierarchies and joint configurations to prepare for Three.js integration.

Comprehensive reference for RFdiffusion3 (RFD3) protein design, covering input parameters, symmetry, hotspots, and best practices for AI-driven protein structure generation.

Executes ML experiments on Modal with error handling, W&B logging, and documentation, supporting collaborative research workflows.

Enables semantic search across past coding agent sessions to retrieve previously solved problems and conversation history.

Provides CRUD operations, schema management, and data import/export for SQLite databases via Python.

Proposes minimal next pull requests for AI governance sessions by analyzing agent configurations and evaluation sets with verification and effort labels.

Context management system for AI conversations, providing TLDR summaries, conversation tracking, and efficient post-compaction recovery.

Automates deduplication and spelling correction for word lists, prioritizing word.txt files to maintain clean and accurate word databases.

Specializes in designing distributed SQL databases with robust schemas, RLS security, and PowerSync for offline-first data synchronization, ensuring continuous availability.

Expert in Mapbox integration for map visualization, location tracking, route planning, and geocoding optimization.

Interactive guide for tuning readability and semantic scoring systems using data-driven experiments to improve obfuscation detection accuracy.

Enables MySQL database connection, schema exploration, and context retrieval for data analysis tasks.

Enables structured knowledge retrieval, context assembly, and semantic anchoring via the Promethean Knowledge Engine framework.

Executes tasks while leveraging machine learning to adapt and optimize performance through iterative learning.

Configures Ralph user settings including AI model selection, profiles, timeouts, and context limits.

Teaches the BAML programming language for generating structured outputs from large language models.

Fetches web content, extracts structured data, takes screenshots, and crawls documentation sites for current information and data extraction needs.

Teaches how to use the edgartools library for accessing SEC EDGAR filings via an API wrapper.

Generates production-ready SSML scripts for Google Cloud Text-to-Speech, enabling precise control over speech synthesis parameters and output.

Creates and refines AI agent personas for Jules using role-specific prompts to ensure precise task execution and project standard compliance.