Data Science & ML Agent Skills

Parxy Document Processing

Parses, converts, and manipulates documents through a unified interface supporting multiple engines like LlamaParse and PyMuPDF.

Pandapower Systems Analysis

Analyzes and optimizes electric power networks using the pandapower Python library.

OpenCL Development

Accelerates parallel computing development using the OpenCL SDK for cross-platform GPU and CPU execution.

CRN Topology & Reaction Networks

Generates and analyzes chemical reaction network (CRN) topologies to predict dynamical behaviors through graph theory and linear algebra.

Infinity Topoi Framework

Implements higher topos theory and ∞-sheaf logic to provide a rigorous mathematical framework for complex systems and BCI data modeling.

CBLAS & LAPACKE API Reference

Provides a comprehensive C API reference for BLAS and LAPACK numerical linear algebra routines.

Dynamic AI Architectures

Implements and manages neural networks that adapt, grow, or prune their topology during training to optimize capacity and prevent catastrophic forgetting.

Rubato Composer Integration

Implements Guerino Mazzola’s mathematical music theory using categorical structures, Scheme integration, and the Topos of Music framework.

World A Topological Interface

Facilitates advanced topological computation and chemputer configuration within the plurigrid ecosystem.

Information Geometry

Implements Fisher-Rao metrics and natural gradient optimization on statistical manifolds for advanced probabilistic modeling.

Curiosity-Driven Learning

Implements Schmidhuber's curiosity-driven framework to provide AI agents with intrinsic motivation based on compression progress.

Active Inference Self-Validation Loop

Validates triadic color system predictions using active inference and error minimization loops.

Fal.ai Audio Mastery

Integrates fal.ai audio models for high-accuracy speech-to-text, premium text-to-speech, and advanced voice cloning.

Unworlding Involution Dynamics

Extracts frame-invariant patterns and self-inverse derivation structures from interaction dynamics using GF(3) conservation.

GGML ML Tensor Library

Optimizes machine learning inference and training using a high-performance C tensor computation library with multi-backend support.

Curiosity-Driven Learning

Implements intrinsic motivation for AI agents using Schmidhuber's formal theory of creativity and compression progress.

AMD ROCm & HIP Development

Develops high-performance GPU kernels and manages AMD ROCm compute stacks using HIP for cross-platform portability.

CUDA GPU Computing

Accelerates high-performance computing tasks by providing expert guidance on NVIDIA CUDA kernels, memory management, and parallel programming libraries.

Three-Match Algorithm Gadget

Facilitates 3-SAT reductions to colored subgraph isomorphism using non-backtracking geodesic gadgets and Möbius filtering.

Agent-o-rama Learning & Pattern Extraction

Trains cognitive agents to extract behavioral, temporal, and network patterns from complex interaction sequences.

Derived Categories & Homological Algebra

Implements advanced homological algebra frameworks including chain complexes, derived functors, and triangulated categories for BCI signal processing.

Emergent Role Assignment

Orchestrates spontaneous role assignment and hierarchical self-organization in multi-agent systems using topological and chemical organization principles.

JaxLife Open-Ended Simulator

Simulates agentic artificial life and emergent behaviors within Turing-complete programmable environments using JAX.

Agent-o-Rama Learning Engine

Extracts complex behavioral patterns and trains predictive learning agents from interaction sequences using multi-interpreter analysis.

Local Feedforward Learning

Implements the Forward-Forward algorithm and local contrastive learning paradigms to train neural networks without backpropagation.

JaxLife Open-Ended Simulation

Implements high-performance agentic simulators using JAX for researching emergent behaviors and open-ended evolution.

Categorical Composition Architecture

Implements mathematical category theory abstractions to enable compositional learning and structured parameter transfer in AI models.

fal.ai Text-to-Image

Generates high-quality AI images using leading models like FLUX, SDXL, and Ideogram through the fal.ai API.

Fal.ai Performance & Cost Optimizer

Optimizes fal.ai AI model deployments for maximum speed, minimal latency, and reduced infrastructure costs.

fal.ai Image-to-Video Animation

Animates static images into high-quality cinematic videos using leading AI models like Kling, Luma, and Runway via the fal.ai API.