3. Data, Science & AI

Expert assistance for PostgreSQL database management, covering schema design, indexing, query optimization, and transaction handling.

Dispatches multiple independent AI agents to concurrently investigate and resolve unrelated system failures without shared state dependencies.

Provides research on Vercel AI SDK patterns, streaming, and multi-provider integration using Exa code search and Ref documentation.

Optimizes PostgreSQL database performance through query tuning, indexing strategies, and execution plan analysis.

Automates PDF image-to-markdown conversion with OCR, structural splitting, layout cleanup, and paragraph-aware translation.

Autonomous agent for initiating ML projects, generating PRDs via conversation, and executing experiments without manual intervention.

Enables creation of custom Claude Code subagents for specialized tasks like code review, debugging, and testing.

Creates standardized .faf context files to provide AI tools with comprehensive project understanding, eliminating repetitive codebase explanations and context drift.

Performs clinical research statistical analysis including survival analysis, propensity score matching, and ROC curve evaluation for treatment effect comparison and study design.

Generates CDISC SDTM/ADaM compliant data schemas from clinical trial protocols for data ingestion and regulatory submissions.

Integrates a new LLM-powered search provider into Mentha, enhancing search capabilities with AI-driven results.

Executes end-to-end data analysis with EDA, descriptive statistics, correlation, and publication-ready Table One reports from CSV datasets.

Designs robust ETL data pipelines with validation using Python libraries like Pandas, Dask, and PySpark for efficient data processing.

Quick reference guide for 51 AutoML MCP tools, listing tool names, functions, and parameters to accelerate machine learning development.

Scans repository skill directories to create a global map of agent capabilities, inputs, and outputs for master agent decision-making in complex task execution.

Automates retrieval, download, and sharing of analysis results and reports from MinIO storage.

Automates ML model training for classification and regression using AutoML, including data processing, evaluation, and prediction workflows.

Automates data cleaning including missing value handling, categorical encoding, feature filtering, and multicollinearity checks for analysis-ready datasets.

Discovers causal relationships, identifies confounders, and constructs DAGs using LLMs for causal modeling and analysis.

Audits SQL migration files for destructive actions, table locks, and compatibility issues to prevent production downtime and ensure data integrity.

Automates discovery and acquisition of computational science datasets including force fields, pseudopotentials, and crystal structures for research workflows.

Executes Quantum ESPRESSO for first-principles quantum mechanical calculations including DFT, SCF, structural relaxation, and band structure analysis.

Queries materials databases (Materials Project, NIST, PubChem) for crystal structures, properties, phase diagrams, and thermodynamic data.

Generates ETL/ELT pipelines for data ingestion, transformation, and warehouse loading.