What the tech

Search

/ opens search from anywhere. Esc closes it.

Tools

Get the newsletter

Skill Finder

The best Claude skills for debugging

Start with systematic-debugging from Jesse Vincent: it is the top skill made for debugging, at 84, needs nothing extra, and forbids Claude from proposing a fix until it has found the root cause. The list also covers Flutter layout errors, InsForge backends, React Native performance, Vue 3 symptoms, an agent push called PUA, and secondary picks for Azure, Supabase, local web apps, Next.js and Android.

The 7 skills made for this job

  1. 1

    systematic-debugging

    For developers whose AI keeps guessing at fixes and who want the root cause found first. Skip it if you just need a typo or one-line slip corrected.

    Score 84 329K installs

    Get systematic-debugging
  2. 2

    flutter-fix-layout-issues

    For Flutter developers stuck on overflow stripes or unbounded constraint errors. Skip it if your bug is in state, data or speed rather than layout.

    Score 80 42.4K installs Official

    Get flutter-fix-layout-issues
  3. 3

    argent-react-native-optimization

    For React Native developers chasing slow screens, jank or extra re-renders who want measurement before changes. Skip it if you do not use the Argent toolkit or your app is not React Native.

    Score 79 30.5K installs Official

    Get argent-react-native-optimization
  4. 4

    insforge-debug

    For developers on InsForge chasing an error, a slow endpoint or a pre-launch audit. Skip it if your backend runs on a different platform.

    Score 78 61.6K installs Official

    Get insforge-debug
  5. 5

    react-native-best-practices

    For React Native developers chasing dropped frames, slow startup, big bundles or memory leaks. Skip it if you are not on React Native or have no measured performance problem yet.

    Score 71 32.1K installs Official High-risk warning

    Get react-native-best-practices
  6. 6

    vue-debug-guides

    For Vue 3 developers chasing runtime errors, warnings, reactivity bugs or hydration mismatches. Skip it if you work in React or need Vue style guidance rather than fixes.

    Score 68 28.1K installs

    Get vue-debug-guides
  7. 7

    PUA

    For developers tired of agents that give up or claim fixes without proof. Skip it if you don't want Claude talking like a hard-driving corporate boss.

    Score 51 19.8K GitHub stars

    Get PUA

Also useful for debugging

Their main job is something else, but they help with this one.

  • Azure skills

    For teams on Microsoft Azure who want Claude to plan, check and run deployments or build Foundry agents. Skip it if you do not use Azure.

    Score 88 687K installs Official

    Get Azure skills
  • supabase

    For developers building on Supabase who want auth, RLS and migrations handled with care. Skip it if your backend runs on plain Postgres or another platform.

    Score 88 363K installs Official

    Get supabase
  • webapp-testing

    For developers who want Claude to click through a web app running on their own machine and see what breaks. Skip it if the site is only live on a remote server.

    Score 86 203K installs Official

    Get webapp-testing
  • next-dev-loop

    For Next.js developers on Turbopack who want each edit confirmed in a real browser. Skip it if you're below Next.js 16.3, on webpack, or not running next dev.

    Score 78 25.2K installs Official

    Get next-dev-loop
  • argent-android-emulator-setup

    For mobile developers using Argent who need Claude to boot and connect to an Android emulator. Skip it if you test on iOS only or don't use Argent's MCP tools.

    Score 76 30.8K installs Official

    Get argent-android-emulator-setup
  • react-doctor

    For React teams who want a health score and regression check before each commit. Skip it if you do not want Claude running npx packages or following a playbook fetched from the web.

    Score 76 21.7K installs Official

    Get react-doctor
  • playwright-best-practices

    For teams writing Playwright end-to-end, component or API tests who want a guide picked for each task. Skip it if your tests use another runner.

    Score 74 105K installs High-risk warning

    Get playwright-best-practices
  • swiftui-expert-skill

    For SwiftUI developers who want help writing, reviewing or profiling views, including Instruments traces. Skip it if you are not building with SwiftUI or cannot run Xcode tools on a Mac.

    Score 73 36.3K installs

    Get swiftui-expert-skill
  • ast-grep

    For developers who need to find code by its structure, like async functions without try-catch. Skip it if plain text search already finds what you need.

    Score 73 19.6K installs Official

    Get ast-grep
  • golang-testing

    For Go developers writing, reviewing or debugging tests, from table-driven cases to fuzzing and leak checks. Skip it if your code is not in Go or you only need testify API details.

    Score 72 45.5K installs

    Get golang-testing
  • fixing-motion-performance

    For front-end developers whose CSS or JS animations stutter or whose scroll effects jank. Skip it if you want new motion designed or a switch to another animation library.

    Score 72 28.8K installs

    Get fixing-motion-performance
  • Android Skills

    For Android developers who want official step-by-step guidance for Compose layout, upgrades, security and performance work. Skip it if you don't build Android apps.

    Score 71 10.5K GitHub stars Official

    Get Android Skills

How to choose

systematic-debugging works in any codebase. Claude reads the full error, reproduces the bug, compares with working code and tests one hypothesis at a time, then writes a failing test before one fix. If three fixes fail, it stops to discuss the design with you.

Pick a framework skill when the error is specific. flutter-fix-layout-issues matches Flutter layout messages, such as RenderFlex overflow, to a known fix and hot reloads until the stripes are gone. vue-debug-guides maps Vue 3 symptoms, from a ref not updating to hydration mismatches, to short reference files.

For React Native speed, react-native-best-practices measures, changes, then measures again and reverts if nothing improved; at least one scanner rated it high risk. argent-react-native-optimization runs a four-phase pass with sub-agents and profilers, and needs the Argent toolkit and a running build.

insforge-debug fits InsForge backends only, reading logs, metrics, database health and RLS policies through the InsForge CLI. PUA, written in Chinese, has Claude speak as a demanding boss when fixes keep failing, and forbids claiming a fix without verification output.

Secondary picks include webapp-testing, which captures console logs from a local web app, Azure skills, supabase, next-dev-loop, react-doctor, Android Skills, and playwright-best-practices, which has a high-risk scanner result.

See other jobs