claude-skills

Implement proven backend architecture patterns including Clean Architecture, Hexagonal Architecture, Domain-Driven Design, and deep module design. Use this skill when designing clean architecture for a new service, when refactoring a monolith to use bounded contexts, when implementing hexagonal or onion architecture patterns, when debugging dependency cycles between application layers, or when making architectural decisions about module depth, seams, and interfaces.

Read from and write to Kyle's Notion Second Brain. Use this skill whenever the user wants to add a note, save an idea, log something, file something away, create a task, write a journal entry, update Notion, or mentions "second brain", "notion", "add to my notes", "save this", "log this", "file this", "add a task", "to do", "remind me", "journal entry", or asks what projects or topics exist. Also trigger when the user shares something worth capturing (research finding, meeting outcome, decision, idea) and asks you to save it. When in doubt, use this skill — it's better to trigger unnecessarily than to miss a capture.

Research-backed learning system with two modes: Study (pre-test → material creation → NotebookLM podcast prompts) and Test (retrieval practice at the edge of your knowledge with gap tracking across sessions). Use this skill whenever the user wants to learn new material, study a topic, create study notes, review for exams, test their knowledge, do retrieval practice, track learning progress, or mentions "study mode", "test mode", "learning", "study", "review", "quiz me", "test me", "knowledge gaps", or "spaced practice". Also trigger when the user uploads PDFs, docs, or other content and says they want to learn or understand it. This skill should be used even for casual requests like "help me learn this" or "I need to study for..."

pytest-specific testing patterns — fixtures, mocking, parameterization, async testing, and configuration. Use when writing Python tests with pytest, setting up pytest infrastructure, or needing pytest-specific patterns. For TDD process and philosophy (red-green-refactor, vertical slicing), use the `tdd` skill instead.

Surgical code refactoring to improve maintainability without changing behavior. Surfaces architectural friction and proposes deepening opportunities — turning shallow modules into deep ones. Covers deep vs shallow module analysis, deletion test, seam identification, extracting functions, breaking down god functions, improving type safety, and eliminating code smells. Use when user says "refactor", "clean up this code", "improve architecture", "make this more testable", or wants to find refactoring opportunities in a codebase.

Skill for analyzing research papers in the context of an active research project. Extracts structured information (contribution, method, results, limitations), assesses relevance to project goals, identifies actionable items, and logs findings to the research log. Use this skill whenever the user shares a paper (PDF, markdown, or link), mentions "analyze this paper", "review this paper", "what's relevant from this", "extract from this paper", "literature review", "paper summary", "how does this relate to our project", "add this to the literature", or provides a research paper with project-related context. Also trigger when the user asks to compare methods across papers, identify gaps relative to existing literature, or update the literature review. Do NOT use for studying/learning a topic (use optimal-learning instead) — this skill is for project-oriented extraction, not personal knowledge building.

Test-driven development with a red-green-refactor loop. Builds features or fixes bugs one vertical slice at a time. Use when user wants to build features or fix bugs using TDD, mentions "red-green-refactor", "test-first", "write tests for this", wants integration tests, or asks for test-driven development. Works with any language — Python, TypeScript, Go, etc. For language-specific test tooling (pytest, vitest, jest), defer to language-specific skills.

Production TypeScript coding standards for modern strict-mode TypeScript (5.x+). Use when writing, reviewing, or refactoring TypeScript to ensure adherence to strict mode, discriminated unions, Zod validation, proper error handling, and modern patterns. Trigger on "make this typesafe", "TypeScript best practices", "improve these types", "strict mode", "type narrowing", "Zod schema", or any TypeScript code quality question. Not framework-specific — applies to React, Node, serverless, CLI, or any TS project.

Apply a visual design theme to a web UI (HTML/CSS/JS, React, Vue, or any frontend). Loads themed design tokens, component patterns, typography, effects, and animation rules from bundled theme files. Use this skill whenever the user wants to restyle, redesign, or improve the visual appearance of a web app or page — especially when they mention a specific aesthetic ("cyberpunk", "minimalist", "noir", "brutalist", "glassmorphism", etc.), share a design system prompt, say a UI looks "ugly", "boring", or "like a kid made it", or ask to apply a theme. Also trigger when the user attaches an HTML/JS/CSS file and wants it re-skinned, or when they want a control panel, dashboard, or tool to look more polished.

Write authoritative white papers for startups — the kind that make investors lean forward, enterprise buyers circulate internally, and founders look like the domain expert they are. Produces structured, research-backed documents that build a case from first principles: defining the problem, analysing why existing solutions fail, presenting the startup's approach as the logical answer, and grounding everything in market evidence and technical depth. Use this skill whenever the user wants to write a white paper, produce a thought leadership document, create a market thesis paper, write a technical white paper, build an investor-facing research document, draft a solution brief, or mentions "white paper", "thought leadership", "market thesis", "position paper", "technical brief", "solution paper", "industry report", or wants to turn startup analysis into a persuasive, outward-facing document. Also trigger when the user says "write something that makes us look credible", "we need a document for enterprise buyers",

Evaluate startups with the rigor of a top-tier VC — covering value proposition, market, team, defensibility, unit economics, technical approach, and website/content quality. Synthesizes classic frameworks (Sequoia, YC, a16z, NFX) with AI-era considerations into structured, actionable reports. Use this skill whenever the user wants to evaluate a startup, review a pitch deck, analyze a company's website or positioning, assess a business model, do due diligence on a startup, review founder materials, critique an adtech/SaaS/marketplace/AI company, or mentions "startup review", "business analysis", "evaluate this company", "what do you think of this startup", "review this pitch", "due diligence", or shares a startup's website asking for feedback. Also trigger when the user wants to write a report for a founder, prepare investor notes, or compare a startup against its competitive landscape.

Annotate AI-assisted reports with authority tiers to signal which claims are the human author's own positions vs. agent-generated analysis. Use this skill whenever the user wants to add credibility annotations to a report, document, or analysis that was co-authored with an AI agent. Also trigger when the user mentions "authority tiers", "claim annotation", "who wrote what", "author vs AI", "human vs agent claims", "confidence tags", "annotation system", "signal my involvement", "which parts did I write", or wants to distinguish their own expertise from AI-generated analysis in any document. Use this even if the user just says something like "I want to show which parts are mine" or "add credibility markers" or "tag the claims".