Skill UI

This meta-skill is designed for complex decision analysis when a user provides or references structured documents such as contracts, quotes, spreadsheets, or notices. Instead of simple summarization, it extracts key evidence, identifies risks (e.g., renewal risk, payment terms), and structures the output into decision-ready formats, such as 'sign/reject/negotiate' recommendations, evidence tables, or required next actions. It supports input from PDF, DOCX, and XLSX formats.

This meta-skill processes an entire Excel workbook to extract deep business insights. It systematically reads all data, performs structured trend and anomaly analysis, writes the comprehensive results back into a dedicated 'Insights' sheet, and permanently stores key performance indicators (KPIs) in long-term memory. Ideal for detailed business review and data deep-dives.

This utility reads structured JSON outputs from multi-search-engine APIs (e.g., Google Scholar, arXiv) provided via standard input. It converts these normalized records into a clean, minimal BibTeX format, suitable for academic papers. It intelligently preserves crucial metadata like DOI, authors, and years while automatically handling data sanitization, deduplication of DOIs, and protecting complex fields like corporate names.

This utility reads the raw content of a UTF-8 text file and emits it verbatim to stdout. It is specifically designed for meta-skills that require round-tripping an artifact through disk, allowing users to manually edit structured files (like scripts or YAML) during review pauses. Unlike standard built-in readers that add line numbering, this method preserves the original bytes, making it suitable for downstream parsers.

Riffkit analyzes the emotional structure, pacing, and successful 'formula' of a winning short video (like a viral TikTok). It then regenerates an entirely new, original video using this proven formula, customized with your product, character, and desired language (English or Spanish). This skill is ideal for creating high-quality, scalable User-Generated Content (UGC) ad creatives or marketing promos for platforms like TikTok and Meta.

A comprehensive tool to scan all connected digital marketing platforms (e.g., Google Ads, Meta, TikTok) for statistically significant deviations from established performance baselines. It detects performance drops (traffic, conversions, ROAS) or unexpected growth opportunities, classifying anomalies by severity (Critical, Warning, Info) and providing probable root causes and immediate action plans. Essential for proactive campaign management and cost control.

Automates the generation and delivery of detailed marketing performance reports from various sources (Google Ads, Meta Ads, GA, etc.). It calculates KPIs, identifies trends, detects anomalies, and generates actionable business recommendations. Reports can be formatted for weekly pulses, monthly reviews, or Quarterly Business Reviews (QBRs), and delivered via Slack, email, or Google Sheets.

A powerful skill designed to analyze raw ad campaign performance data from platforms like Google, Meta, and LinkedIn. It goes beyond simple metrics reporting, providing deep diagnosis by identifying underperforming ads, pinpointing budget waste, and calculating the true Customer Acquisition Cost (CAC). It generates concrete, actionable recommendations for what to cut, scale, or test next, optimizing your ad spend and maximizing ROAS.

A comprehensive guide for authors preparing their accepted KDD paper for ACM proceedings. It covers the entire production pipeline, including page budgeting (9 content pages, 3 pages for references/appendix), managing e-rights and TAPS uploads, de-anonymization steps, and ensuring proper Digital Library metadata for final publication.

This advanced skill specializes in detecting and recovering hidden data (steganography) embedded within various media containers, including images, audio, video, and documents. It is crucial for forensic analysis and CTF challenges, allowing users to inspect metadata, LSBs, alpha/palette channels, and appended trailers to reconstruct hidden payloads without brute force.

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.

This robust fixture is designed to execute predefined tasks using a specified seed path and current working directory. It intelligently separates dispatch metadata within YAML frontmatter from the main content body, ensuring the router reliably loads execution parameters while ignoring potential interference from the body text. Ideal for structured testing and controlled execution workflows.