Use when authoring TypeScript libraries or npm packages - covers project setup, package.json exports, build tooling (tsdown/unbuild), API design patterns, type inference tricks, testing, and publishing to npm. Use when bundling, configuring dual CJS/ESM output, or setting up release workflows.
100
Does it follow best practices?
Validation for skill structure
New library project scaffolding
pnpm scripts
60%
100%
prepublishOnly script
100%
100%
ESM-first type:module
100%
100%
Dual CJS/ESM exports
57%
100%
Types in .d.mts
0%
100%
sideEffects false
0%
100%
files dist only
100%
100%
tsconfig moduleResolution Bundler
100%
100%
isolatedDeclarations true
0%
100%
verbatimModuleSyntax true
0%
100%
noEmit true
0%
100%
tsdown dual format
100%
100%
tsdown dts and clean
100%
100%
antfu ESLint config
0%
100%
antfu type lib
0%
100%
vitest test location
100%
100%
Without context: $0.3451 · 1m 17s · 18 turns · 67 in / 4,359 out tokens
With context: $0.7363 · 1m 36s · 35 turns · 31 in / 5,239 out tokens
Package exports and build multi-entry
Three export entries
100%
100%
Types .d.mts extension
0%
100%
Types condition first
0%
100%
Dual CJS/ESM per entry
75%
100%
sideEffects false
100%
100%
files dist
100%
100%
Legacy main/module/types fields
87%
100%
typesVersions fallback
0%
100%
tsdown format esm+cjs
100%
100%
tsdown dts enabled
100%
100%
tsdown three entries
100%
100%
type module
100%
100%
package validation note
0%
0%
Without context: $0.2732 · 1m 3s · 13 turns · 13 in / 4,001 out tokens
With context: $0.3953 · 1m 14s · 21 turns · 17 in / 3,976 out tokens
CI and release automation
pnpm/action-setup@v4
0%
100%
setup-node cache pnpm
0%
100%
CI separate jobs
100%
100%
CI node version 22
0%
100%
Release node version 24
0%
100%
OIDC permission id-token write
100%
100%
publish with --provenance
100%
100%
wait-for-ci job
42%
100%
changelogithub in release
0%
100%
repository field in package.json
100%
100%
bumpp release script
0%
100%
changelogen in release script
0%
100%
No NPM_TOKEN
100%
100%
concurrency cancel
0%
100%
Without context: $0.4626 · 2m 27s · 13 turns · 14 in / 9,301 out tokens
With context: $0.5701 · 1m 32s · 27 turns · 156 in / 5,354 out tokens
Table of Contents
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.