Data Science & ML Agent Skills

Topological Gesture & Hypergesture Modeler

Models continuous performance curves and high-order musical gestures using topological category theory and Mazzola's Diamond Conjecture.

Polars Portfolio Backtester

Performs ultra-fast portfolio backtesting and trading strategy analysis using the Polars library.

LLM Performance Evaluation

Implements comprehensive evaluation frameworks for LLM applications using automated metrics, human feedback, and benchmarking strategies.

LangChain App Architect

Builds sophisticated LLM applications using the LangChain framework with advanced agent, memory, and tool integration patterns.

World Hopping Logic

Navigates complex mathematical and ontological possibility spaces using Badiou-inspired event logic and Kripke semantics.

Rerank - Search Result Optimization

Enhances search precision by applying cross-encoder models to re-order and refine initial vector search results.

Swarm Orchestration for Multi-Agent Systems

Orchestrates multi-agent swarms using advanced coordination patterns and dynamic topologies for complex, parallel task execution.

ReasoningBank with AgentDB

Implements high-performance adaptive learning and experience replay for self-improving AI agents using AgentDB.

Qualitative Document Converter

Automates high-fidelity PDF and document conversion by selecting the optimal parsing method for academic and qualitative research data.

Qualitative Research Methodological Rules

Ensures methodological rigor in qualitative research by enforcing isolation rules and phase-aware guidance through specialized AI agents.

Qualitative Coding Workflow

Orchestrates systematic document coding, progress tracking, and audit trail generation for qualitative research projects.

Qualitative Interview Ingestion

Converts audio recordings and PDF documents into structured markdown for qualitative research and analysis.

Qualitative Reflection & Reasoning

Facilitates deep analytical reasoning and navigates theoretical paradoxes in qualitative research using systematic thinking and dialectical wisdom.

ML Pipeline Workflow

Orchestrates end-to-end MLOps pipelines from data preparation through production deployment and monitoring.

Qualitative Data Ingestion

Converts audio recordings, PDFs, and diverse document formats into structured markdown for qualitative research and AI-assisted analysis.

Qualitative Research Project Setup

Initializes qualitative research environments with structured epistemic foundations and standardized folder hierarchies.

Qualitative Analysis Orchestrator

Optimizes model selection, API cost estimation, and batch processing strategies for AI-assisted qualitative document analysis.

LangChain Architecture Design

Architects sophisticated LLM applications using LangChain patterns for autonomous agents, stateful memory management, and modular chains.

Qual-Shared Research Infrastructure

Manages the foundational state, logging, and workflow enforcement for AI-assisted qualitative research projects.

Zeroth Bot Robotics Platform

Facilitates the development, simulation, and control of 3D-printed humanoid robots for reinforcement learning research.

Prompt Engineering Patterns

Optimizes LLM performance and reliability through advanced prompting techniques like chain-of-thought and few-shot learning.

Qual Analysis Orchestrator

Orchestrates cost-effective qualitative research workflows using Kimi K2.5 models and advanced context caching strategies.

MuJoCo Scene Builder

Composes complex 3D environments and terrains for robotic simulation and reinforcement learning training.

Entropy Sim2Real Bridge

Bridges the gap between robotic simulations and real-world deployment using maximum entropy reinforcement learning and information-theoretic alignment.

Kinfer Runtime

Deploys trained Reinforcement Learning (RL) policies to real robots using high-performance Rust and ONNX Runtime.

RAG Implementation Expert

Builds Retrieval-Augmented Generation (RAG) systems to ground LLM applications with vector databases and semantic search capabilities.

Qual-Coding: Qualitative Research Partner

Facilitates reflexive qualitative data analysis through dialogical questioning and multi-stage visible reasoning.

Deep Analytical Reasoning

Facilitates structured, step-by-step thinking for complex analytical decisions and theoretical framework construction using the Sequential Thinking MCP.

K-Bot Humanoid Robotics

Provides hardware specifications, MuJoCo MJCF models, and deployment workflows for the K-Scale flagship humanoid robot platform.

K-Scale Robotics Ecosystem

Provides a unified framework for humanoid robot development, reinforcement learning training, and sim-to-real deployment.