Skill UI

Analyzes accumulated ARIS usage logs to propose optimizations for SKILL.md prompts, reviewer settings, and workflow defaults. A read-only producer that stages patches for human-gated review.

This skill involves creating structured specifications before writing any code to serve as a shared source of truth between developers and engineers. It is used when starting new projects, features, or significant changes where requirements are unclear or ambiguous. The workflow includes a scope check, specifying objectives, planning, task breakdown, and implementation with human reviews at each phase. It helps decompose complex requirements into capability maps and ensures clarity before coding begins.

This skill provides comprehensive management for annotation schemas (configs) and human review workflows (queues) within the Arize platform. It supports defining categorical, continuous, and freeform label types. Users can programmatically apply these annotations to project spans and datasets using the Python SDK, facilitating structured human feedback loops and improving model monitoring.

Three-layer verification pipeline that extracts verifiable claims from AI output, searches the web for supporting or contradicting sources, flags hallucinations, and builds structured reports for human review.

This skill prepares and validates research-only clinical decision-support artifacts, including evidence profiles, cohort tables, and survival analysis plans. It supports aggregate or synthetic research documentation and traceability. It excludes patient care and live clinical operations, focusing on governance, privacy, and statistical evaluation for qualified human review.

This expert skill audits and designs user interfaces for the Apple ecosystem (iOS, macOS, watchOS, visionOS) against the official Human Interface Guidelines (HIG). It checks crucial elements like contrast ratios (WCAG standards), minimum tap-target sizes, and adherence to modern design languages like Liquid Glass. Use it when reviewing mockups or developing native-feeling Apple apps to ensure accessibility and optimal user experience.

Scores inbound deals on margin and risk, routes discount requests to a named human approver chain, and flags contract landmines in redlined MSAs. Never auto-approves; every output is a scorecard plus routing recommendation.

Render ARIS Markdown or JSON artifacts such as IDEA_REPORT, AUTO_REVIEW, KILL_ARGUMENT and research-wiki state into polished single-file HTML reports for human reading, with sticky TOC, math, callouts, score progressions and source drift detection.

This meta-skill automates the complex process of modernizing legacy PDF documents (e.g., historical contracts, scanned manuals, legal documents). It follows a structured four-step pipeline: structural extraction, natural-language rewriting of problem pages, generating an audit summary, and finally re-merging the content into a single, refined PDF. This ensures content is modernized while maintaining fidelity to the original meaning, providing a human review summary before finalization.

A human-in-the-loop studio for iteratively tuning AI outputs or reviewing complex media/prose. It allows users to directly manipulate visual parameters (color, typography, layout) or annotate documents in real-time, moving beyond static comparison grids or vague prompts. Perfect for designers and content creators who need hands-on, visual feedback.

This skill structures a highly controlled, bounded development life cycle for AI agents. It mandates a structured three-phase workflow: Planning (Spec), Implementation (Build), and Verification (Review). It is essential for building complex features from scratch, especially when dealing with ambiguous requirements, high risk, or needing explicit human approval gates to ensure the resulting code meets all defined specifications.

Rewrites overly formal or vague code review comments generated by AI or corporate tools. It cuts out filler phrases like "I noticed..." or "It might be worth considering...", transforming them into direct, actionable feedback. The output uses a clear format (Location: Severity: Observation. Fix.) that sounds like it came from a helpful, experienced teammate, significantly improving the quality and reception of PR feedback.