Skill UI

Guides users through a structured 23-step interview to design, generate, and deploy Claude plugins across claude.ai and Claude Code, handling activation prompts, platform quirks, and GitHub pushes.

This skill acts as a rigorous, Socratic interviewer designed to stress-test a proposed plan against an established domain model and existing project documentation. It forces the user to sharpen terminology, resolve cross-cutting dependencies, and identify inconsistencies between stated requirements, code implementations, and documented decisions. It ensures the resulting architecture is precise, consistent, and well-documented by automatically updating the context and generating Architecture Decision Records (ADRs).

This skill provides a comprehensive, multi-phase guide for authors to draft a Specification Enhancement Proposal (SEP) adhering to rigorous governance standards. It walks users through prerequisites, initial vetting (Gate), detailed interviews (e.g., breaking changes, security), and technical research, ensuring the proposed change is viable and community-vetted before formal submission.

This skill guides developers through creating a comprehensive refactoring plan and serving as a Request for Comments (RFC). It orchestrates detailed interviews, codebase analysis, and scope definition. It ensures the refactor is broken down into tiny, safe, incremental commits, culminating in a structured GitHub issue that serves as a definitive architectural and technical decision document.

Stress-test design plans against project documentation including CONTEXT.md glossaries and ADR decisions. Interview users one question at a time to resolve terminology and decisions. Explore codebase to verify claims. Create ADRs only when decisions are hard to reverse, surprising without context, and result from real trade-offs. Maintain documentation consistency throughout the session.

This tool extracts and standardizes comprehensive paper-writing preferences from a user's initial request. It determines whether the generation should proceed directly using safe academic defaults or if it needs to initiate a preference interview, ensuring all structured data required for robust academic drafting is captured.

This skill guides you through a comprehensive, multi-phase interview to capture every necessary detail of a new project idea. It is designed to run BEFORE any code is written, forcing detailed specification gathering across purpose, features, data models, and technical constraints. The output is a clean, self-contained prompt.md file, ensuring subsequent development is based on solid, unambiguous requirements and preventing costly mistakes.

This tool generates a comprehensive hiring rubric and technical interview scorecard for evaluating software engineers at specific levels and roles. It defines clear behavioral anchors, structured technical questions, and system design criteria, ensuring consistent, unbiased evaluation across all interviewers, and providing a full debrief agenda.

A comprehensive, structured guide for designing large-scale distributed systems. Use this framework to structure interview answers or real-world architectural proposals, covering everything from clarifying requirements and estimating capacity to deep component design, data flow, and critical trade-offs.

This skill guides users through an in-depth interview process to achieve maximum project context completeness (100% "Gold Code") for their .faf file. By systematically asking questions about project goals, architecture, and technology stack, it ensures all critical context slots are filled, making the project fully structured and highly readable by AI systems.

Scan and manage a registry of existing git repositories and linked worktrees, set default repos as interview context, and rank repos by relevance to a topic using lexical BM25 search.

Establishes project quality standards as a written contract to prevent agents from lowering them. Interviews users on key dimensions, sets sane defaults, and records constraints in CONSTRAINTS.md. Monitors for weakened bars like suppressed checks or skipped tests. Use when defining quality gates, setting coverage thresholds, or ensuring autonomous agents adhere to standards.