Skill UI

A powerful Python library for analyzing, validating, converting, and transforming complex materials structures and associated computational data. It provides functionalities for handling compositions, periodic crystals, molecules, local phase diagrams, symmetry assignments, and electronic structure I/O, ensuring strict provenance tracking for all transformations.

Draft, revise, and audit scientific manuscripts and reports with rigorous evidence provenance, authorship accountability, and confidentiality controls. This tool ensures every factual claim is traceable to verified evidence, making it essential for high-stakes academic writing, journal submissions, and regulated research reporting where accuracy and compliance are non-negotiable.

This skill facilitates reproducible retrieval of scientific, regulatory, or financial facts by querying documented public APIs. It requires defining a strict retrieval contract, handling complex filtering, managing pagination, and ensuring full provenance and auditable results, minimizing reliance on general knowledge inference.

This skill integrates 10 major academic literature APIs (including PubMed, arXiv, OpenAlex, Crossref, etc.). It guides users on converting complex research intents into reproducible, verifiable lookups for papers, preprints, citations, and open-access full texts. It ensures results include verifiable provenance and follows a rigorous workflow for maximum accuracy in scholarly research.

This tool functions as a post-task epilogue, designed to audit and record the entire research journey. It scans the conversation history at the end of a coding or research session to extract critical research artifacts, including decisions, experiments, dead ends, claims, heuristics, and major pivots. It systematically writes this provenance data into the structured `ara/` directory, ensuring a faithful and auditable trace of how the project truly evolved.

MockHunter is an advanced skill designed to audit live web pages, providing a comprehensive reality check on visible data. It analyzes every element to determine its true source: whether it is REAL (from the database), MOCK (simulated data), LLM-generated, HARDCODED, BROKEN (broken endpoints), or UNKNOWN. Ideal for auditing AI-generated UIs or reviewing proof-of-concept MVPs before customer sign-off.

Provides high-level strategic guidance for founders and Chief Data Officers (CDOs) in startups. This skill covers critical decisions such as AI training data rights and consent provenance, selecting optimal data architectures (warehouse, lakehouse, data mesh), valuing customer data as a strategic asset for M&A or fundraising, and planning data team organizational evolution. It focuses exclusively on strategic decision-making, not tactical engineering.

This skill provides deep, decision-grade research dossiers on any specific entity (company, person, or organization). Unlike generic profiles, it mandates the user to state a clear hypothesis upfront. The resulting dossier rigorously tests this hypothesis, providing a conclusive verdict, detailed activity timelines, network signals, red flags, and a source-provenance audit log. Ideal for due diligence, investment analysis, or high-stakes meeting preparation.

A sophisticated pipeline that generates professional, print-ready academic conference posters (e.g., for ICLR, NeurIPS). It builds the poster as a single HTML/CSS file and renders it precisely to PDF using headless Chromium. The system enforces measurement-driven hard gates, utilizes a two-hue design system, and ensures figure provenance, guaranteeing structural integrity for publication standards.

This skill enables the creation and maintenance of highly structured, configurable research wikis. It manages source provenance, compiling raw source materials into organized concept pages, index maps, and derived artifacts. Ideal for systematically documenting complex knowledge bases, such as research notes, product documentation, or domain studies, ensuring full traceability and consistency.

This skill provides comprehensive tools to detect data poisoning and model backdooring attacks throughout the machine learning lifecycle. It checks data provenance, detects label-flip anomalies (using Cleanlab), and analyzes model activations and spectra (using ART) to identify malicious triggers or compromised weights, ensuring ML supply chain integrity.

This skill guides the end-to-end process of supply chain security by generating standards-compliant Software Bills of Materials (SBOMs) in CycloneDX and SPDX formats. It details using Syft to create inventories, correlating them with vulnerability intelligence via Grype, enforcing CI/CD gates based on severity, and establishing verifiable provenance using Cosign signing. Essential for modern DevSecOps practices.