3. Data, Science & AI

Expert assistance for designing, building, debugging, and integrating Model Context Protocol (MCP) servers in LLM applications.

Creates isolated subagents for specialized AI tasks, enabling modular and context-specific execution within larger AI systems.

Guides users in developing agent skills for the Photon AI agent platform, enabling effective AI agent functionality and workflows.

Creates persistent project rules for AI guidance within Photon, ensuring consistent AI behavior and project-specific adherence.

Enables real-time text-to-speech with on-device voice synthesis and cloning on Apple Silicon, using MLX for private, no-API-key audio generation.

Enables episodic memory retrieval, pattern analysis, and troubleshooting for the Memory MCP server, including validation of MCP implementations.

Enables complex task execution through intelligent planning and coordination of multiple specialized agents with dynamic optimization and quality gates.

Guides users in creating, configuring, and refining AI agents through persona definition, system prompt generation, and specialized skill assembly for diverse applications.

Creates and manages user-defined rules to customize AI behavior for personalized interactions and project-specific configurations.

Guide for integrating Vercel AI SDK into Remix apps to build AI chatbots, text completion, and agents for Shopify.

Analyzes ClickHouse schema design, partitioning, ORDER BY keys, and materialized views to identify anti-patterns and optimize table structures for performance.

Analyzes ClickHouse SELECT query performance to identify slow patterns and provide optimization strategies for reduced latency and timeouts.

Diagnoses and tracks ClickHouse mutation operations (ALTER UPDATE, DELETE) to resolve stuck mutations and performance issues.

Analyzes ClickHouse cache systems (mark, uncompressed, query) to resolve cache hit ratio issues and optimize database performance.

Diagnoses ClickHouse merge performance, part backlog, and 'too many parts' errors to optimize database operations and prevent failures.

Analyzes ClickHouse query failures, exceptions, and error patterns to diagnose and resolve database issues efficiently.

Diagnoses ClickHouse storage performance including disk usage, compression efficiency, and IO bottlenecks to optimize database resource utilization.

Diagnoses ClickHouse memory usage, OOM errors, and memory pressure to resolve out-of-memory issues.

Diagnoses and optimizes ClickHouse data ingestion performance, addressing slow inserts and pipeline bottlenecks through batch sizing and part creation analysis.

Formulates data migration plans and outlines secure procedures for transferring data between systems.

Guides execution of evaluation workflows for AI model benchmarking, including scoring methodology and pipeline management for SF-Bench.

Validates database integrity and referential consistency to ensure data reliability after modifications or before committing changes.

Provides real-time statistics on the Reality Check database, including counts for claims, sources, chains, and predictions.

Provides offline text-to-speech synthesis using the sherpa-onnx framework, enabling local processing without cloud dependency.