3. Data, Science & AI

Enables management of multi-agent orchestration within Claude's MPM framework for complex AI workflow coordination.

Manages configuration settings for Claude's Model Parameter Management (MPM) system, enabling efficient model setup and tuning.

Adjusts prompts, models, and tool configurations for the review-item-processor AI agent to refine its functionality.

Guides AI feature integration in CodeSpirit, including LLM implementation, prompt engineering, and AI-driven business functionality development.

Expert assistance for writing, optimizing, and debugging GROQ queries used in Sanity.io data retrieval and manipulation.

Configures search, filters, scopes, and sorting for Plutonium resource queries to enable efficient data retrieval and management.

Generates weekly analytics and project activity summaries for Claude Code usage, including statistics on projects worked on and code activity.

Configures OpenCode CLI settings including default model, providers, and baseURL for AI model integration.

Orchestrates Gemini CLI agent execution via tmux for parallel coding tasks, overcoming headless mode limitations with 1M+ context window.

Generates detailed analytical summaries by fetching and processing data from external APIs for user requests.

Guides editing of master data YAML schemas for adding, modifying, and deleting schema elements and structures.

Adds and configures custom AI models (self-hosted, non-standard providers) by researching API specs and updating configuration files.

Guides users through creating and optimizing AI agents for software development via interactive configuration and performance tuning.

Creates and refines OpenCode AI agents via guided Q&A, handling configuration, permissions, and performance optimization.

Guides optimal task delegation to subagents in AI workflows for enhanced problem-solving efficiency.

Provides a command-line interface for Google NotebookLM, enabling programmatic management of notebooks, sources, and AI-generated content like reports and podcasts.

Automatically informs Claude that it operates in a restricted environment with only the execute_code tool available.

Generates MongoDB schema definitions in YAML for ezorm, including collections, fields, indexes, and relations without Go code output.

Provides A-share value investment analysis including stock screening, deep analysis, industry comparison, and valuation using public financial data.

Optimizes MLX model performance and memory usage through compilation tuning, lazy evaluation, and memory management for faster execution.

Adds a new tool to the agent's toolstore registry for LLM-based agent applications.

Creates a voice-based AI agent using YAML configuration and customizable prompt templates for LLM interactions.

Creates new evaluation scenarios to test and validate the performance of AI agents in controlled environments.

Executes Instrumental Variables, Difference-in-Differences, and Regression Discontinuity Design analyses in R with diagnostic validation.