3. Data, Science & AI

Provides optimized patterns for ClickHouse schema design, MergeTree engine selection, and query performance tuning to handle slow queries and batch inserts.

Enables real-time data synchronization from relational databases to ClickHouse using CDC methods, handling schema changes and data consistency.

Handles continuous stream ingestion from Kafka, RabbitMQ, or Kinesis into ClickHouse with backpressure, exactly-once semantics, and performance optimization.

Provides CLI access to Model Context Protocol (MCP) servers for connecting to external AI tools, APIs, and data sources.

Automatically selects and composes the optimal AI skill for user requests without explicit specification, using intelligent fallback strategies.

Guides creation and configuration of AI agents in DEVS platform, including behavior modification and handling agent YAML/JSON configuration files.

Guide for integrating LLM providers in the DEVS platform to add LLM functionality and create AI-powered features.

Generates professional performance analysis reports from SDL3 HammerEngine benchmark data, featuring statistical analysis, visualizations, and optimization recommendations for documentation and release reporting.

Enables real-time web search using Google's Gemini AI model to fetch current information upon user request.

Guides creation, refinement, and optimization of specialized AI agents including model selection, focus definition, and tool configuration.

Validates AI agent configurations including model selection, tool permissions, and approach quality to ensure optimal performance and best practices.

Enables semantic similarity-based vector search on Agentuity cloud platform with authentication for AI-driven data retrieval.

Executes authenticated SQL queries on Agentuity cloud database for platform operations management.

Adds or updates vectors in vector storage for Agentuity cloud platform operations, requiring authentication for secure AI data management.

Retrieves specific vector entries by key from Agentuity cloud platform, requiring authentication for secure data access.

Generates SQL schema for authentication tables to manage user credential storage and access control.

Evaluates machine learning models for performance, fairness, and reliability using metrics like accuracy, precision, and cross-validation.

Optimizes prompts for large language models, covering system prompts, agent signals, and reasoning techniques to enhance AI agent performance and security.

Creates data visualizations, dashboards, and reports for analytics, ML, and infrastructure monitoring using appropriate libraries and accessibility principles.

Analyzes and optimizes SQL queries for performance, including index design, query rewriting, and execution plan analysis with PostgreSQL-specific tuning.

Analyzes THOR's SQLite database to identify performance bottlenecks, track scan timing, and compare deltas for debugging and optimization.

Designs optimized database schemas, data models, and relationships for relational, NoSQL, and data warehouse systems with normalization, indexing, and ETL strategies.

Enforces strict SQL schema conventions for PSRDatabase to ensure consistency and maintainability in database design.

Automates session lifecycle management by inheriting prior knowledge and depositing key insights for context-aware AI interactions.