Skill UI

A comprehensive framework for rigorously testing and benchmarking LLM agents. It moves beyond simple pass/fail checks to assess complex behaviors, reliability metrics, and capability consistency across multiple runs. Ideal for production monitoring and identifying subtle failure modes in advanced AI agents.

A sophisticated hybrid memory system designed specifically for AI agents. It provides a persistent, searchable knowledge bank to store, retrieve, and manage long-term knowledge, including architectural designs, operational patterns, and critical decision points. By syncing with project documentation, it ensures that crucial institutional knowledge remains accessible throughout the development lifecycle, enhancing agent reliability and consistency.

A comprehensive reference architecture for implementing large-scale AI image generation using Ideogram. This system guides developers through building robust pipelines, including structured prompt templating (for brand consistency), utilizing multiple API endpoints (Generate, Edit, Remix), and implementing full asset workflow management from generation to final CDN delivery. Ideal for establishing brand asset systems and scalable AI integrations.

This skill provides methodologies for designing sophisticated, layered memory architectures for AI agents. It moves beyond simple context windows and basic vector stores by incorporating short-term, long-term, and graph-based memory systems. Implementing this is crucial for agents that require persistence across sessions, maintaining complex entity consistency, and performing advanced, relationship-driven reasoning over accumulated knowledge bases, including time-aware fact tracking.

This skill formalizes successful, recurring patterns discovered in auto-memory (MEMORY.md) and graduates them into permanent, enforced rules within the project's core instruction set (CLAUDE.md or scoped rules). It ensures that learned best practices become mandatory project conventions, significantly improving adherence and consistency across all future AI interactions and development cycles. Use this when a temporary note proves to be a vital, repeatable guideline.

Autonomously improve a generated paper through LLM review, fix implementation, and recompilation over two rounds. Use it to polish theory consistency, soften overclaims, and fix presentation after a paper is compiled.

A comprehensive scoring tool to validate localized or translated content before publishing. It calculates a composite score by analyzing technical accuracy (e.g., formatting preservation, placeholder integrity, length ratios), overall content quality, brand voice consistency, and adherence to target market compliance rules. Use this to replace subjective checks with structured, repeatable quality assurance.

This comprehensive guide assists researchers in converting an accepted, anonymous draft for the Conference on Learning Theory (COLT) into the final, polished camera-ready version for publication in PMLR proceedings. It covers critical steps such as de-anonymization (restoring author details, acknowledgements), consolidating all promised revisions into the main text, verifying PMLR metadata and BibTeX consistency, and preparing the full archival PDF.

A comprehensive guide for authors submitting to PR&AI, detailing how to truthfully declare and provide verifiable artifacts (code, data, models) for peer review. It emphasizes honesty, consistency, and structuring a reproducible repository to enhance reviewer trust.

A comprehensive guide for creating the two required artifacts for a WACV paper: an anonymous package for double-blind review, and a public release post-acceptance. It emphasizes reproducing deployed system claims (latency, power, robustness) and maintaining consistency across revision rounds.

A comprehensive guide for strengthening the reproducibility of academic papers, especially those involving systems claims (e.g., power usage, latency). It details maintaining a 'recipe ledger' for all reported metrics, ensuring constraint-aware measurement of environmental variables, and managing data/artifact consistency across the entire two-round revision process.

This meta-skill automates the entire short-drama production pipeline, starting from a simple topic. The workflow analyzes the request to infer render style, character identity, and shot count. It first drafts a detailed, shot-by-shot shooting script and pauses for user review. After approval, it generates a universal cast reference image, per-shot composition images, and corresponding video clips. Crucially, it maintains character and scene consistency by anchoring every video to both the reference image and the composition image. Finally, it adds title/ending cards, burns subtitles, and de...