3. Data, Science & AI

Displays user skill inventory, stat sheet, or skill tree visualization upon trigger.

Orchestrates Task Decomposition Protocol v2 for task decomposition (/plan) and execution via subagents, delegating state management to tasker CLI.

Orchestrates parallel execution of multiple AI agents in multi-agent workflows, managing independent task streams.

Automates system learning from user interactions to enhance future assistance, triggered after productive sessions and problem-solving.

Provides database schema design, migration creation, SQL query optimization, and operations for PostgreSQL, Supabase, and Prisma.

Manages persistent storage and retrieval of conversational facts and user preferences for context-aware AI interactions.

Enables complex cross-document queries via knowledge graph construction and stateful multi-hop reasoning for large-scale document analysis beyond standard RAG.

Designs scalable and performant database schemas for SQL and NoSQL databases, including normalization, indexing, and migration strategies.

Builds AI-native applications with RAG pipelines, vector databases, agentic workflows, and LLM integration, including prompt engineering and cost optimization.

Enforces strict SQL schema conventions for the Quiver project to maintain consistent database structure and design.

Enables collaborative creation and iterative refinement of AI agent skills using evaluation-driven development cycles.

Manages Model Context Protocol (MCP) servers for AI agents, enabling extension of agent capabilities through new tools and data sources via configurable transports and scopes.

Enables semantic recall of past conversations and historical context using FTS5 and Chroma to enhance response relevance.

Provides real-time access to cryptocurrency wallet balances, transaction history, and associated addresses.

Enables collaborative refinement and stress-testing of outcomes by consulting external experts via Codex MCP, utilizing AI for expert-level insights.

Intelligently extracts structured data from web pages and sites using Firecrawl, enabling efficient data collection for analysis and processing.

Tool for designing relational database schemas, including table structures, entity relationships, and index optimization for MySQL and PostgreSQL.

Creates comprehensive Databricks lakehouse pipeline design documents with bronze/silver/gold layers, SLAs, data quality, and governance for ETL/ELT, DLT, and batch/streaming jobs.

Optimizes Databricks SQL queries for performance, correctness, and readability by analyzing joins, filters, aggregations, and partition pruning.

Automates academic research by scraping web sources, generating BibTeX bibliographies, and providing templates for scientific writing with citation management.

Queries Apple Health SQLite database for vitals, activity, sleep, and workouts, outputting in Markdown, JSON, or FHIR R4 for health analytics.

Command-line tool for processing text and multimedia files via multiple LLM providers with model configuration and file handling.

Analyzes meeting transcripts to extract decisions, action items, opinions, questions, and terminology using AI.

Queries and organizes browsing history across synced devices using natural language and LLM-based categorization, outputting to markdown or JSON.