3. Data, Science & AI

Coordinates multiple AI agents through structured ritual-inspired patterns for synchronized multi-agent execution.

Filters test/dummy revenue data from real customer transactions across all revenue streams including affiliates, gallery, and subscriptions.

Provides command-line interaction with EVM blockchains for querying data, sending transactions, and calling smart contracts via Cast CLI.

Imports data from SQLite databases or shared chitta files for data migration workflows.

Extracts text content from PDF files using pdfplumber for data processing and analysis tasks.

Checks database schema integrity and data consistency, ensuring critical columns and legacy data migrations are valid.

Creates and optimizes Claude Rules configuration files in markdown format, following best practices for AI behavior definition.

Assists in writing PostgreSQL SQL queries, creating Node.js models with pg, debugging database errors, and analyzing table structures.

Enables building AI agents with features defined in prompts instead of code, supporting autonomous agents, MCP server design, and self-modifying systems.

Sends prompts to remote Claude AI instances on VPS for distributed collaboration, supporting multiple model backends and independent contexts.

Optimizes AI agent systems for cross-environment compatibility (CLI, IDE, GitHub) and integrates emerging AI technologies for enhanced performance.

Enables semantic search for conversation memories, allowing AI to recall past interactions and provide context-aware responses.

Explores MongoDB database structures and data for debugging, investigation, and understanding game data patterns.

Enables semantic search to recall and utilize past conversation context, enhancing interaction continuity and relevance.

Creates Claude skills using Anthropic best practices to package procedural knowledge and extend AI domain expertise through structured capabilities.

Creates custom D3 visualizations with unique interactions for non-standard questions, serving as a fallback solution.

Provides agent execution patterns for LLM calls, token handling, and vision processing in AI agent frameworks.

Enables calling Foundry agents to rate tasks, prioritize work, and interact with the ratemytask agent.

Generates DuckDB SQL queries for data analysis on CSV, Parquet, and .ddb files, including query optimization and setup assistance.

Provides instant data file previews, schema inspection, and simple filtering for CSV and JSON files.

Fetches minimal, task-relevant context to enhance AI model responses without overwhelming the system.

Provides best practices for crafting system prompts in agentic workflows, optimizing agent behavior and output for modern language models.

Creates executable Python functions for matplotlib/seaborn visualizations from natural language plot specifications.

Enables SQL-based analysis of CSV, JSON, Parquet, and Excel data files for filtering, aggregation, and exploration.