3. Data, Science & AI

Provides data handling best practices for Guidewire InsuranceSuite, covering entity management, migration, batch operations, and governance.

Configures local development environments for Databricks using dbx and Databricks Connect, enabling efficient testing and iteration for data engineering workflows.

Generates and edits images using Google Gemini API through the nanobanana platform.

Creates optimized meta-prompts for Claude-to-Claude pipelines with research, planning, and execution stages to enable sequential AI output consumption.

Guides creation, configuration, and usage of Claude Code subagents and the Task tool for specialized AI agent workflows.

Guides creation of Multi-Agent Systems with centralized (TeamSwarm) or decentralized (Swarm workflow) coordination, setting up agent structures via filesystem operations.

Optimizes AI model workflows through competitive code compression, enhancing efficiency and reducing data overhead in MCP contexts.

Loads CausalPy's internal example datasets for demonstration and testing in causal inference projects.

Enables interactive development and execution validation in Marimo notebooks for data science and AI workflows.

Performs causal inference analysis using CausalPy for DiD, ITS, SC, and RD methods to estimate treatment impacts.

Diagnoses and optimizes AI skill execution issues such as context explosion and agent coordination failures using Gemini CLI for deep analysis.

Transforms vague user prompts into clear, actionable specifications using AI analysis and session context. Activated by -e or --enhance flags.

Graph analysis tool for Beads projects, computing PageRank, betweenness, critical path, and cycles. Supports AI agent integration via --robot-* flags.

Builds automated machine learning pipelines with code generation, data validation, performance metrics, and documented artifacts for ML workflows.

Automates SQL transformation tasks for data pipeline workflows, enabling efficient data manipulation and processing.

Analyzes database query performance using EXPLAIN plans to identify bottlenecks and provide optimization recommendations for faster execution.

Provides model interpretability via SHAP and LIME for debugging, fairness assessment, and stakeholder communication.

Classifies text emotional tone into positive, negative, or neutral for customer reviews, social media, and surveys.

Generates Luigi task operations for data pipeline orchestration. Auto-activates for seamless data processing workflow setup.

Real-time cryptocurrency portfolio tracker with valuation, allocation analysis, and P&L calculations.

Automates learning rate scheduler configuration and management for machine learning training pipelines.

Automates feature engineering tasks for machine learning models, enhancing data preparation efficiency during training.

Generates flame graphs for performance profiling, visualizing function call stacks and CPU usage in applications.

Analyzes blockchain gas price patterns to optimize transaction costs and timing for cost-effective transactions.