Skill UI

This skill guides the design and implementation of a complete, production-grade Machine Learning pipeline. It employs a multi-agent orchestration workflow, systematically covering all stages of the ML lifecycle: data ingestion and quality assurance, feature engineering, model training, optimization, and robust deployment (MLOps). It ensures built-in reproducibility, monitoring, and scalability.

Mastering the full lifecycle of machine learning models from data ingestion to deployment. This expertise covers building robust, production-grade MLOps pipelines using orchestrators like Kubeflow, Airflow, or Prefect. Key skills include feature store implementation with Feast, rigorous data validation, experiment tracking (e.g., MLflow), model registry management, and automated deployment workflows. Essential for automating model training and ensuring reproducibility at scale.

Pyvene is a declarative framework designed for performing causal interventions on PyTorch neural networks. It allows researchers to conduct advanced experiments such as activation patching, causal tracing (ROME-style), and interchange intervention training (IIT). Use this library when you need to test causal hypotheses about model behavior, deeply interpret model components, or ensure reproducibility in advanced AI research.

LaminDB is an open-source lineage-native lakehouse designed for biological datasets and models. It enables queryable, traceable, and validated data management across various storage formats like S3 and Postgres. Users can track computational workflows, curate data with ontologies via Bionty, and ensure FAIR compliance. Ideal for managing scRNA-seq, spatial transcriptomics, and integrating ML pipelines while maintaining reproducibility and collaboration.

Assists authorized reviewers in preparing evidence-bounded, constructive peer-review drafts and structured manuscript assessments. Supports reporting-guideline selection, claim–evidence checks, methods, statistics, reproducibility, and ethics critique.

The BIDS skill provides a comprehensive framework for organizing, standardizing, and managing complex neuroscience and biomedical datasets. It covers querying, validating, and converting various modalities (MRI, EEG, MEG, etc.) to ensure compliance with the BIDS standard for research reproducibility and efficient data sharing.

This skill provides a rigorous, measured framework for optimizing system performance. It guides users to convert vague goals (e.g., 'make it faster') into a bounded, repeatable process. The loop requires establishing a clear baseline, systematically testing hypotheses (variants), identifying bottlenecks, and promoting the safest, most measurably improved solution, ensuring reproducibility and stability throughout the process.

Guides users through the entire end-to-end process of running the GAIA benchmark. It covers environment validation, cost estimation, model execution, and generating a signed, HAL-compatible submission package for leaderboards. This ensures high integrity and reproducibility for model performance evaluation.

A comprehensive tool for creating and updating highly specific audience segments across multiple platforms (CRM, Email, Ad Networks). Users can define segments using advanced criteria, including RFM modeling, behavioral data (page views, purchases), demographic details, and engagement status. This ensures campaigns are targeted precisely, supports cross-platform synchronization, and maintains clear documentation for reproducibility.

A comprehensive guide for preparing scholarly manuscripts involving spatial data. It outlines best practices for documenting data provenance (sources, projections, processing chains), ensuring reproducibility, and adhering to critical ethical standards such as geoprivacy and human-subjects protection. Essential for maintaining research integrity in academic publishing.

This comprehensive guide provides authors with best practices for preparing supplementary artifacts for AAAI submissions. It outlines how to structure technical appendices, package reproducible code and data ZIPs, maintain strict anonymity (by removing author metadata and paths), and maximize the evidence for reproducibility. The goal is to ensure that external reviewers can easily verify the research claims without specialized or private tools.

Provides comprehensive guidelines for designing, auditing, and documenting empirical evidence for high-level AI research submissions (e.g., AAAI). It covers essential methodologies including strong baselines, single-factor ablations, robustness testing (shift, seed, prompt), statistical significance, and ensuring perfect alignment with reproducibility checklists and ethical reporting. Use this guide before submission to ensure empirical evidence strongly supports all claimed AI contributions.