REPO PACKAGE
mcollina skills
COMMUNITY11skills
6ready to use
LABSCO SUMMARY
Three skills form a coherent Node.js/Fastify core: node covers native TypeScript type-stripping, async patterns, and error handling on Node 22+; fastify-best-practices covers routes, plugins, JSON Schema validation, and the Pino-backed request lifecycle; and oauth layers OAuth 2.0/2.1 flows — PKCE, device flow, token introspection — on top of a Fastify app specifically. nodejs-core is narrower still: it's for someone drafting or reviewing a commit against the nodejs/node repository itself, down to V8 deoptimizations and node-gyp builds, not for an application developer.
The rest are general-purpose developer skills with no shared thread: documentation applies the Diátaxis framework to any project's docs, init writes or prunes a repository's AGENTS.md, linting-neostandard-eslint9 migrates a project to ESLint v9 flat config, typescript-magician handles generic types and any removal, octocat drives git and GitHub operations from any pasted GitHub URL, skill-optimizer tunes other Claude skills for activation reliability, and snipgrapher turns source files into syntax-highlighted PNG, SVG, or WebP snippet images.
READ THE FULL ANALYSIS
Who gets the most out of it. Anyone building a Fastify backend gets three skills that reinforce each other (node, fastify-best-practices, oauth); anyone who isn't on Fastify still gets seven general-purpose skills with no Fastify dependency, and one — nodejs-core — that's useful only if you're contributing to Node.js itself. Six of the eleven run with nothing extra at all, four assume some local CLI is already on the machine (ESLint, the gh CLI, snipgrapher's own tool), and oauth is the one classed as needing real credentials to exercise, since it's built around testing actual token flows.
WHAT'S INSIDE
11 showing · 11 totalReady to use
Nothing else to set up — install it and go.
11 of 11
SETUP
Showing all 11 skills
documentation
Documentation gets confusing when a lesson, a set of instructions, a list of facts and a background explanation all land on the same page; this works out which of the four you are writing, and shapes it to suit.
fastify-best-practices
Fastify is a framework for building web servers in JavaScript, and this is the rulebook for building one properly — laying out the routes, checking what callers send in, logging, testing and going live.
init
Most of what an AI coding assistant needs about a codebase it can read for itself; this writes the short note covering everything left over — the local habits, the odd commands and the traps it would never guess.
linting-neostandard-eslint9
Sets up the automatic checker that flags style slips and likely mistakes in JavaScript code, and moves a project that is still on the older, retired way of configuring it across to the current one.
node
Practical Node.js know-how for code that survives production — running TypeScript with no build step, handling errors, streaming files too big for memory, chasing down flaky tests, profiling slow paths, and shutting down without dropping work.
nodejs-core
Node.js itself is a large C++ program, and this is about changing that program rather than something built with it — finding the crash, rebuilding, and submitting the patch the way its maintainers expect.
oauth
Adds the kind of login where people sign in through an outside provider — Google, GitHub, a company account — to a web app built with Fastify, along with the checks that stop a stolen sign-in being reused.
octocat
Git and GitHub work done by typing commands instead of clicking through the website, with a checked, step-by-step route through the operations that are easiest to get wrong.
skill-optimizer
When an AI model keeps ignoring the instructions you wrote for it, this measures how often that happens, rewrites the wording until it sticks, and re-runs the test to prove the change helped.
snipgrapher
Turns a piece of code into a good-looking picture of it — the kind people post on social media or drop into documentation — using a command-line tool called snipgrapher.
typescript-magician
TypeScript catches bugs by knowing the shape of your data; this fills in the places where that knowledge is missing or was deliberately switched off, until the compiler reports nothing wrong.
No skill in this package matches that.
HOW TO GET IT
WHOLE PACKAGEInstall the whole package (11 skills)
npx skills add mcollina/skillsSINGLE SKILLOr install a single skill
npx skills add mcollina/skills --skill <name> --full-depthPick the skill name from the Skills tab — each entry there installs independently.