Skill UI

This skill transforms a course syllabus (PDF, DOCX, etc.) into a highly curated, supplementary academic reading list. It parses topics and learning outcomes, calibrates content difficulty based on the target audience (undergraduate, graduate, etc.), and searches academic databases (Consensus) for recent, peer-reviewed papers. The output is a professional .docx file containing summaries, links, and advanced discussion questions, making it invaluable for educators and students needing structured, up-to-date course resources.

This skill autonomously optimizes artifacts like code or models using Hypothesis Tree Refinement. It runs iterative experiments with dev/test evaluators to prevent overfitting, coordinating subagents in isolated worktrees. Ideal for improving training recipes, agents, or pipelines without manual supervision.

This skill executes the DeepSWE benchmark, a comprehensive coding-agent evaluation suite. It uses OpenRouter and mini-swe-agent to score various Large Language Models (LLMs) on 113 tasks, providing reproducible results for model comparison, development assessment, and research benchmarking. Use this when a standardized, deep evaluation of LLM coding capabilities is required.

This framework guides AI agents to critically assess their own output across five key axes: Accuracy, Completeness, Clarity, Actionability, and Conciseness. It provides a structured, evidence-based scorecard and concrete suggestions for improvement, turning every task completion into a deep learning opportunity.

This pattern defines a robust, continuous loop structure for autonomous agents. It integrates crucial components like quality gates, evaluation harnesses, and explicit recovery controls. It guides complex workflows by defining decision paths (e.g., PR control, RFC decomposition) and includes strategies to mitigate common failure modes such as infinite loop churn or cost drift, ensuring measurable progress in autonomous systems.

This skill provides a detailed, side-by-side comparison of advanced LLM agent evaluation frameworks, such as ruflo and HAL. It helps researchers and developers identify capability gaps—including real code execution, full web browsing, and structured file parsing—and prioritize architectural improvements for robust agent benchmarking and performance tracking.

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 sophisticated multi-agent development harness inspired by GANs. It separates product planning, code generation, and rigorous evaluation into three distinct agents (Planner, Generator, Evaluator). This creates an adversarial feedback loop, allowing projects to iteratively reach professional, production-grade quality far beyond what a single-agent workflow can achieve. Ideal for complex, full-stack applications.

This skill facilitates the crucial process of sharing learned intelligent patterns across isolated projects or machines using the InterPlanetary File System (IPFS). It allows users to publish a project's accumulated knowledge base to a decentralized network and later fetch, apply, and merge these patterns into new or existing projects. This is invaluable for bootstrapping new applications with prior knowledge or consolidating learnings across a large monorepo.

Provides a structured workflow for machine learning engineering tasks, covering model development, training, evaluation, and deployment. Use it to manage end-to-end ML projects efficiently.

A comprehensive guide detailing the entire model lifecycle for RuView. It supports multiple advanced tracking methods including camera-free pose estimation (WiFlow), camera-supervised methods using MediaPipe, robust contrastive embeddings (RuVector), and domain generalization (MERIDIAN). Covers training, evaluation, and deployment across local GPU and cloud platforms.

This tool allows users to fine-tune the evaluation criteria for AI-generated content. Users can configure minimum scoring thresholds, adjust the relative weights of dimensions (such as brand voice, hallucination risk, and claim verification), and set composite auto-reject rules. It ensures the quality bar matches specific brand guidelines, industry needs, or content formats, making quality trade-offs explicit and adjustable.