Skill UI

This skill analyzes an existing codebase using domain context and architectural decisions to identify architectural friction and "deepening opportunities." It proposes refactored improvements to increase module depth, enhancing testability, improving maintainer locality, and making the codebase more navigable for AI tools. It focuses on transforming shallow modules into robust, deeply leveraged components.

A comprehensive, step-by-step guide for responding to Klaviyo API outages and integration failures. This runbook includes quick triage scripts, decision trees based on common error codes (401, 429, 5xx), and detailed procedures for mitigation, service degradation, and postmortem communication, ensuring rapid service recovery and operational stability.

A tool designed for persistent context management within AI coding sessions. Memory Kit solves the problem of context loss that occurs during compaction or session restarts by saving, loading, updating, and sharing critical session state to a dedicated `MEMORY.md` file. Key functions include creating snapshots of goals and decisions, syncing context across team members via Git, and auditing/pruning stale entries, ensuring continuous workflow and knowledge retention.

A comprehensive runbook for rapidly responding to Notion API incidents. It provides structured triage flows, automated health checks, and a decision tree to classify failures (Notion-side vs. integration-side). Use it for immediate diagnosis of outages, error mitigation (e.g., rate limits, token expiry), and post-incident reviews.

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.

A comprehensive guide for rapid incident response when Shopify integrations fail or behave unexpectedly. It instructs users to triage issues by checking the status page, diagnosing API connectivity, and handling common errors like rate limits (429) or expired tokens (401). Includes decision trees, immediate mitigation steps, and communication templates.

This comprehensive guide serves as a decision flowchart for developers using Claude Code. It details when to use specific slash commands (/plan, /compact, /loop), select specialized agents (planner, code-reviewer, build-error-resolver), or invoke workflow skills (e2e-testing, tdd-workflow). It helps users select the optimal tool combination for complex tasks across the entire software development lifecycle, including feature planning, bug fixing, testing, and architectural design.

A deep-dive session designed to stress-test a technical plan against an existing domain model. This process forces the user to resolve ambiguities, sharpen terminology, and validate decisions by cross-referencing code and existing documentation (CONTEXT.md, ADRs). It ensures architectural rigor and completeness before implementation.

A structured methodology for validating non-trivial decisions by subjecting them to an adversarial, fresh-context review. This process is crucial when correctness is paramount—such as in production, security-sensitive logic, or complex architecture—by forcing a skeptical examination that aims to disprove assumptions rather than simply approve them.

This tool simulates a Chief Technology Officer's comprehensive review, stress-testing any major technical plan. It systematically interrogates critical areas including scaling limits (scaling cliffs), accumulated technical debt, team capacity planning, build-vs-buy economics, system reliability (SLOs), and security compliance. Use it before committing to core architecture changes or making high-stakes engineering decisions.

This tool simulates a rigorous, relentless interview process designed to stress-test any plan, design, or architectural concept. It forces the user to walk through every decision branch, resolving dependencies one by one until a comprehensive, shared understanding is achieved. Ideal for developers and product managers needing to validate complex system designs or roadmaps.

Simulates a comprehensive, multi-stage academic journal peer review process. This agent dynamically configures five expert reviewers (EIC, 3 peers, and a Devil's Advocate) to evaluate a paper across methodology, domain expertise, cross-disciplinary impact, and core arguments. It provides structured reports, editorial decisions, and revision roadmaps, supporting full reviews, post-revision verification, and calibration modes.