NeVibe · 42
Skills for Claude Code, Codex and Cursor
SKILL.md folders the agent loads on its own when a task matches the description.
frontend-design This skill guides intentional visual design for new interfaces and redesigns. Use it when a generic generated UI is not enough.
skill-creator Creates and refines skills, then tests them with prompts, evaluations, and performance checks.
improve-codebase-architecture This skill scans a codebase for architectural friction, then builds a visual HTML report with refactoring candidates. Use it before a deeper architecture review.
tdd Guides the red-green TDD loop: ensures tests are written first, keep them at public interfaces, and avoid coupling to implementation details.
setup-matt-pocock-skills Sets up a repository for Matt Pocock engineering skills with issue tracking, triage labels, and domain-document rules. Run it before using the other engineering skills.
handoff Creates a handoff document from the current conversation so another agent can continue the work. Use it when a task moves to a new session or agent.
triage This skill moves issues and external PRs through triage states, verification, and agent briefs. It fits maintainers managing an issue queue.
prototype Creates throwaway prototypes for checking state logic or exploring UI before committing the decision to the main codebase.
grilling Stress-test a plan, decision, or idea through structured rounds of questions. Use it when you need to expose assumptions before acting.
domain-modeling This skill sharpens project terminology, maintains GLOSSARY.md, and records selected architecture decisions in ADRs. Use it while the domain model is changing.
teach Turns the current directory into a learning workspace and keeps progress across sessions. Use it when learning needs to support a concrete goal in your project.
codebase-design This skill helps design modules with small interfaces and hidden complexity. Use it when refactoring code, choosing a seam, or making tests easier to write.
pptx Handles `.pptx` and `.potx` files for creating, reading, editing, and validating presentations. Use it whenever an agent needs to touch slide files or templates.
diagnosing-bugs This skill guides an agent through reproducible diagnosis of hard bugs and performance regressions. Use it when code fails, produces wrong results, or becomes slow.
code-review Reviews changes from a chosen reference point against repository standards and the originating spec. Use it for a branch, PR, or work in progress.
implement Implements work from a specification or tickets, following a test-driven flow and ending with review and a commit. Use it when a coding task needs a defined delivery routine.
ask-matt Routes you to the right Matt Pocock skill or flow for the task at hand. Use it when you need a starting point for a feature, prototype, incoming issue, or hard bug.
pdf This skill lets an agent read, create, and edit PDF files, including forms and scanned documents. Use it when PDF work is part of the task.
docx This skill creates, reads, and edits Word documents and templates. Use it for reports, letters, and files that need headings, tables, or page numbers.
wayfinder Maps large, uncertain work as decision tickets in an issue tracker. Use it when the route to a goal needs more than one agent session to clarify.
to-spec Turns the current conversation and repository context into a spec, then publishes it to the project issue tracker. Use it when a feature discussion is ready to hand off to an agent.
research Researches a question through primary sources and saves the findings as a Markdown file in the repository. Use it for API facts, documentation checks, and delegated reading.
to-tickets Turns a plan, spec, or conversation into linked tickets with blocking dependencies. Use it when work needs to be handed to agents or published to a tracker.
xlsx Creates, reads, edits, and converts spreadsheet files. Use it when the requested deliverable is an Excel workbook or another tabular file.
webapp-testing This skill tests local web apps with Playwright, helping agents debug UI behavior, capture screenshots, and inspect browser logs.
wizard This skill creates an interactive bash wizard for manual setup steps, CI secrets, third-party dashboards, and one-off migrations.
git-guardrails-claude-code This skill adds a Claude Code hook that stops risky Git commands before they run. Use it when an agent must not push or discard working-tree changes.
to-questionnaire Creates a Markdown questionnaire for someone who holds context you need. Use it when a colleague or client must answer before you can make a decision.
setup-pre-commit This skill configures pre-commit checks for a JavaScript repository, running Prettier, type checking, and tests before each commit.
scaffold-exercises This skill builds course exercise directories with sections, problems, solutions, and explainers. Use it when starting a new course section or creating exercise stubs.
migrate-to-shoehorn Replaces `as` type assertions in tests with helpers from `@total-typescript/shoehorn`. Use it when tests need partial objects or deliberately invalid data.
writing-for-agents Guidance for writing skills and agent-facing documents. Use it when editing `AGENTS.md`, `CLAUDE.md`, or reference files reached from them.
mcp-builder This skill walks through building an MCP server in TypeScript or Python, from reading the protocol spec to writing tests and evaluation questions.
canvas-design Builds posters and visual art in .png and .pdf by first writing a design philosophy, then expressing it on the canvas.
brand-guidelines Applies Anthropic's official brand colors and fonts to documents, slides and other artifacts. Useful whenever something needs to match the company look.
theme-factory Applies preset color palettes and font pairs to slides, documents, and HTML pages. Use it when an artifact needs one consistent visual style.
doc-coauthoring Guides document drafting through context collection, section work, and a reader check. Use it for specs, RFCs, proposals, and decision documents.
algorithmic-art Creates p5.js generative art with seeded randomness and adjustable parameters. Use it for flow fields, particle systems, and code-based visual experiments.
writing-plans Creates an implementation plan from a spec before code changes begin. Use it when a task has multiple steps, files, and tests.
implement-spec Implements a prepared spec from its ticket graph on one integration branch, using parallel subagents. Use it when tickets have blocking dependencies.
1–40 of 42
A skill is a folder with a SKILL.md file: name and description in the frontmatter, instructions below, optional scripts and reference files next to it. The agent keeps only the name and description in context (roughly 100 tokens per skill) and reads the body when a task matches. That is why you can install dozens of them without eating the context window.
The format is an open standard (agentskills.io). Claude Code, Codex, Cursor, Copilot, Gemini CLI, Kiro, OpenCode and Zed all read it; what differs is the folder. Claude Code uses .claude/skills/<name>/ in a project and ~/.claude/skills/ for every project, Codex uses .agents/skills/, and Cursor reads both. Install by copying the folder, through a plugin with /plugin install, or with npx skills add owner/repo -a claude-code.
Step by step: Claude Code, Codex, Cursor. Writing your own: how to write a SKILL.md.
