skills/dev-npxify/SKILL.md
Audit project dependencies and replace CLI-only tools with npx/pnpm dlx to reduce installed packages. Use when: (1) User wants to reduce dependencies, (2) User says 'npxify', 'dlxify', 'reduce deps', (3) User wants to clean up package.json, (4) User asks which deps can use npx/pnpm dlx.
npx skillsauth add takazudo/claude-resources dev-npxifyInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Audit project dependencies and identify packages that can be replaced with npx or pnpm dlx to reduce the installed dependency count.
Check the project for lockfiles to determine the package manager:
pnpm-lock.yaml → use pnpm dlxpackage-lock.json → use npxyarn.lock → use npx (yarn dlx also works but npx is more universal)Find all package.json files in the project (root + workspaces). For each, catalog all dependencies and devDependencies.
For each dependency, determine how it is used:
grep for require('pkg'), from 'pkg', import 'pkg' in source codescripts in package.json for the package binary nameClassify into one of these categories:
These are CLI tools that run infrequently and have few sub-dependencies:
| Package | Typical Usage | Notes |
|---|---|---|
| husky | "prepare": "husky" | One-time setup on install |
| lint-staged | Called from git hooks | Runs only on commit |
| serve / http-server | "serve": "serve dist" | Ad-hoc static file serving |
| create-* / init-* | Project scaffolding | One-time use |
| madge | Dependency graph | Occasional analysis |
| depcheck | Unused dep detection | Occasional analysis |
| license-checker | License audit | Occasional audit |
| npm-check-updates / ncu | Update checking | Occasional use |
| @takazudo/mdx-formatter | Markdown formatting | Infrequent, lint-staged |
| concurrently / npm-run-all | Script runners | If used only in one script |
| rimraf / del-cli | Cross-platform rm | Simple CLI, few deps |
| cross-env | Cross-platform env vars | Simple CLI, few deps |
| dotenv-cli | Env file loading | Simple CLI |
These must stay as installed dependencies:
| Category | Examples | Reason | |---|---|---| | Runtime libraries | react, express, lodash, axios | Imported in source code | | Build toolchains | webpack, vite, esbuild, next, typescript | Deep integration, used constantly | | Framework packages | @docusaurus/, @nestjs/, gatsby | Core framework | | Type definitions | @types/* | Needed for IDE and typecheck | | Test frameworks | jest, vitest, mocha, playwright | Run frequently, deep integration | | Linter + plugins | eslint, eslint-plugin-*, prettier (when used by eslint-plugin-prettier) | Plugins must resolve from same node_modules | | PostCSS/Tailwind | tailwindcss, postcss, autoprefixer | Build-time integration | | Heavy CLI tools (100+ deps) | electron-builder (300+ deps), webpack-cli | Too slow to download each time | | Packages used by other deps | prettier (if eslint-plugin-prettier is installed) | Must be resolvable | | Config dependencies | globals (if imported in eslint.config.mjs) | Imported at config load time |
Present a table to the user:
## Replaceable with pnpm dlx
| Package | Location | Current Usage | Sub-deps |
|---|---|---|---|
| husky | root devDep | prepare script | ~1 |
| lint-staged | doc devDep | pre-commit hook | ~30 |
## Keep as dependency (cannot replace)
| Package | Location | Reason |
|---|---|---|
| eslint | doc devDep | Plugins need local resolution |
| typescript | blog devDep | Build toolchain, IDE integration |
For each approved replacement:
pnpm dlx <package> or npx <package>dependencies or devDependenciespnpm install (or npm install) to update the lockfileWhen unsure whether a package is replaceable, check these:
pnpm why <package> or check npm. If 100+, keep as dep.require/imported, it MUST stay as a dependency.pnpm dlx, BUT if eslint-plugin-prettier is installed, prettier must stay as a devDependency (the plugin requires it to be locally resolvable).pnpm dlx downloads all of them every time — extremely slow. Always keep as devDep.lint-staged config may reference pnpm exec <tool> for tools that ARE installed. Only the lint-staged binary itself can be dlx'd; the tools it invokes still need to be installed.development
Link Claude Code skill names mentioned in a CodeGrid article (data/{series}/{n}.md) to the author's public claude-resources repo, pinned to the latest commit hash so links don't rot. Use when: (1) user says 'linkify cc resources', 'link the skills', 'link skill names', or invokes /dev-linkify-cc-resources; (2) editing a CodeGrid article that mentions `/commits`, `/pr-complete`, `/skill-creator` or other Claude Code skills and they should point to claude-resources. Only links skills that actually exist in the public repo; skips hypothetical examples and code blocks.
development
Second opinion from Claude Opus on a plan or approach. Use when: (1) Planning phase of /big-plan needs a higher-quality review than /codex-2nd / /gco-2nd, (2) User says 'opus 2nd' or 'opus opinion', (3) Wanting Anthropic's larger model to critique a plan. Spawns a general-purpose Agent with model: opus that reads the plan file and returns structured feedback. Anthropic quota — not free.
tools
AI-based testing via subagent + a per-task test-flow skill. Use when the user wants to verify something that mechanical assertions can't fully capture — image recognition, visual size/position comparison, animation smoothness, multi-step manual flows that need AI judgment. Triggers: 'AI-based test', 'AI test', 'visual verify', 'image recognition test', 'manual operation test', 'human-eye check', 'verify visually', 'compare screenshots', 'looks the same', 'looks correct'. The skill's job is to (1) author a focused test-flow skill that captures the exact procedure + verdict criteria, then (2) dispatch a verification subagent via the Agent tool that loads BOTH the test-flow skill AND a browser-driving skill (/verify-ui primary, /headless-browser fallback) so the subagent has clear context and consistent verdicts. NEVER uses `claude -p` — subagent dispatch goes through the Agent tool exclusively.
development
End-of-workflow audit of touched GitHub issues, PRs, and branches via a Sonnet subagent. Use when: (1) /big-plan, /x-as-pr, or /x-wt-teams finishes its main work and needs to verify every touched resource is in the right state (closed when done, kept when ongoing, deleted when dead), (2) User says 'cleanup resources', 'audit cleanup', or 'check what should be closed', (3) A long workflow ends and the manager wants a structured paper trail of what it closed/kept/deleted. Auto-execute by default — the Sonnet agent proposes, the manager (you) executes safe actions and prints a final report.