skills/dev-package-json/SKILL.md
Organize and maintain package.json and npm config (.npmrc) for readability and security. Use when: (1) Reorganizing scripts section or adding separators, (2) Extracting multi-process commands into shell scripts, (3) Setting up multi-environment dev commands (local/preview/prod), (4) Handling pnpm "Ignored build scripts" warnings, (5) Configuring .npmrc security (strictDepBuilds, allowBuilds, ignoredBuilds), (6) Managing pnpm via corepack and packageManager field, (7) Adding predev port cleanup. Keywords: package.json, npm scripts, .npmrc, pnpm, build scripts, supply chain, corepack, packageManager, predev, kill port, port in use.
npx skillsauth add takazudo/claude-resources dev-package-jsonInstall 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.
Two techniques for keeping large scripts sections readable and maintainable.
Add visual section dividers using unused JSON keys:
{
"scripts": {
"// ── Core ─────────────────────────────────────────": "",
"dev": "next dev",
"build": "next build",
"// ── Testing ─────────────────────────────────────": "",
"test": "jest",
"test:e2e": "playwright test"
}
}
Format: "// ── Section Name ──────..." with ─ padding to ~50 chars, value "".
Add a predev script that kills stale processes on dev server ports before starting. This prevents "port already in use" errors that commonly occur after crashes, orphaned processes, or forgotten terminal sessions.
{
"scripts": {
"predev": "lsof -ti :5173,:8787 | xargs kill 2>/dev/null; true",
"dev": "next dev"
}
}
How it works:
lsof -ti :PORT — finds PIDs listening on specified ports (-t = terse/PID-only, -i = internet addresses):5173,:8787 checks multiple ports at oncexargs kill — sends SIGTERM (graceful) to found processes2>/dev/null; true — silently succeeds when no processes are foundpredev before dev (lifecycle hook convention)Adapt port numbers to match your project's dev servers (e.g., :3000,:8080 for a typical Node.js + API setup).
When a command starts 2+ background processes, extract to scripts/*.sh.
Template available at scripts/multi-process-dev.sh.template. Key pattern:
#!/bin/bash
set -e
cleanup() {
kill $PID_1 $PID_2 2>/dev/null
wait $PID_1 $PID_2 2>/dev/null
}
trap cleanup EXIT INT TERM
if [ "$MODE" = "local" ]; then
backend-server &
PID_1=$!
sleep 3
fi
pnpm dev &
PID_2=$!
wait
Call from package.json: "dev:full": "MODE=local ./scripts/dev-full.sh"
Make executable: chmod +x scripts/*.sh
For apps with local/preview/production API targets:
{
"// ── Dev with API (3 environments) ───────────────": "",
"dev:full": "API_MODE=local ./scripts/dev-full.sh",
"dev:full:preview": "API_MODE=preview pnpm dev",
"dev:full:prod": "API_MODE=production pnpm dev"
}
See references/patterns.md for:
_ prefix)packageManager FieldPin the exact pnpm version in package.json:
{
"packageManager": "[email protected]+sha512.36cdc707e7b..."
}
This ensures every developer and CI uses the identical pnpm version. Corepack reads this field and auto-downloads the specified version.
corepack enable
After this, running pnpm install / pnpm dev / etc. just works — corepack intercepts the pnpm command and uses the pinned version automatically. No global pnpm install needed.
pnpm self-update — it errors when pnpm is managed by corepackcorepack use pnpm@latest routinely — it bumps the version in package.json and often regenerates pnpm-lock.yaml, creating noisy diffscorepack use pnpm@<version>, commits the package.json + lockfile changes, and everyone else gets it via pnpm installAI tools and automation sometimes add corepack use pnpm@latest to setup steps. This causes unnecessary version bumps and lockfile churn. Remove it — corepack enable + the existing packageManager field is sufficient.
Evaluate and manage dependency build scripts for supply chain security.
See references/npmrc-build-scripts.md for:
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 / /gcoc-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.