Skill UI

This utility is designed to reconcile the ADR (Architecture Decision Record) index when source files have been deleted or modified, preventing the accumulation of stale or dangling reference rows. Unlike simple upserts, this tool performs a hard purge of the entire index namespace (adr-patterns and adr-edges) and then fully rebuilds the index from the files currently present on disk. It is crucial for maintaining data integrity after deletions.

Profiles a paper and matches it to suitable academic journals from an index of 743 venues, ranking candidates into reach, match, and safe categories with a resubmission ladder. Ideal for authors deciding where to submit or resubmit after rejection.

A comprehensive guide for packaging and submitting artifacts for Information Retrieval (IR) conferences (like SIGIR). It details the strict requirements for reproducibility, including mandatory components like TREC-format run files, qrels, and index recipes. The process covers repository structure, anonymization for review, handling data provenance, and ensuring that results can be audited by any reviewer using standard community tooling.

This comprehensive guide details how to architect a lengthy, complex academic paper into a single, highly navigable PDF. It establishes a two-tier document model, separating self-contained 'Front Matter' from verifiable 'Back Matter.' Learn best practices for structuring proofs, ensuring dependency order, and implementing critical navigation aids (like claim indexes and TOCs) so reviewers can follow the argument flow seamlessly, even without traditional supplementary files.

This skill scans a specified local directory, extracts essential metadata from the files, and builds a comprehensive, searchable index. It then securely uploads this generated index to a configured remote search service via HTTP, making local files discoverable across remote platforms.

A comprehensive guide for creating and maintaining an atomic, index-driven knowledge base (KB) using Markdown. This method enables AI agents to bypass traditional vector search limitations by reading a central index file (INDEX.md) and opening only the most relevant, highly signaled notes, drastically improving context efficiency.

This skill automatically scans all defined user and project skills (from SKILL.md files) to create a comprehensive, centralized index. It manages skill paths, triggers, descriptions, and scopes, ensuring that downstream agents and delegators always have an up-to-date source of truth for available functionalities. It is crucial for maintaining system integrity and avoiding redundant skill definitions.

This skill compiles durable, non-obvious findings into an interlinked markdown knowledge store using atomic files and wiki links. It ensures agents retain knowledge across sessions without relearning facts. It involves creating one fact per file with metadata like confidence and provenance, maintaining an index for easy retrieval. Ideal for research, competitive intel, or investigation results.

Deploy and manage the ELK Stack (Elasticsearch, Logstash, Kibana) for centralized log aggregation, analysis, and visualization. Includes Docker Compose setup, index templates, lifecycle management, Logstash pipelines, Filebeat configuration, and Elasticsearch queries.

This skill enables users to add and manage structured evaluation results within Hugging Face model cards. It supports extracting evaluation tables from README files, importing benchmark scores via the Artificial Analysis API, and executing custom model evaluations using vLLM or lighteval backends. It ensures compatibility with the model-index metadata format for standardized model releases.

Removes Lovable.dev scaffolding from Vite + React projects including dependencies, branding, placeholder assets, and generated documentation. Helps developers ship clean, ownable codebases after exporting from Lovable. Audits package.json, vite.config.ts, index.html, and source files for Lovable fingerprints.