0xbigboss/claude-code
27 skills13.7k installs1.4k starsGitHub
Install
npx skills add https://github.com/0xbigboss/claude-codeSkills in this repo
1React Best Practicesreact-best-practices is a 0xbigboss/claude-code skill that guides React component work with a decision tree for when Effects are appropriate versus event handlers, render-time calculation, useMemo, key-based resets, refs, and useEffectEvent. The core principle treats Effects as escape hatches for synchronizing with external systems such as WebSocket, IntersectionObserver, third-party libraries, and window listeners, not for derived state or user-event responses. It documents anti-patterns including storing derived values in useEffect, effect chains, notifying parents from effects, and reading refs during render. Custom hook guidance prefers focused useXxx hooks that call other hooks, avoids lifecycle helper hooks, and keeps one concrete use case per hook. Component patterns cover controlled versus uncontrolled state, composition over prop drilling, compound components with provider-scoped state, and flushSync when synchronous DOM reads follow updates. The skill pairs with typescript-best-practices for typed React code and defers detailed examples to react-patterns.md. Developers reach for it when reviewing or authoring components that risk redundant effects and extra renders.3kinstalls2Typescript Best Practicestypescript-best-practices is an agent skill for TypeScript and JavaScript files covering type-first patterns, functional style, and runtime validation aligned with project CLAUDE.md conventions. It pairs with react-best-practices whenever work touches React components, tsx files, or React imports because this skill covers language fundamentals only. Core patterns make illegal states unrepresentable through discriminated unions instead of overlapping boolean flags, branded types for domain primitives like UserId, const assertions for literal role unions, and exhaustive switch checks with never. Runtime validation guidance centers on Zod schemas as the single source of truth with z.infer types, safeParse for user input, parse at trust boundaries, schema composition via extend pick omit merge, and transform for normalization at parse time. Optional type-fest utilities include Opaque for branded types, PartialDeep, ReadonlyDeep, SetRequired, SetOptional, and Simplify for complex intersections. Developers reach for it when reading or writing ts, tsx, js files or tsconfig.json and need compile-time safety plus Zod validation patterns aligned with repository standards.3kinstalls3Python Best PracticesUse when reading or writing Python files py pyproject toml requirements txt The python-best-practices skill documents workflows and patterns from the repository SKILL md name python-best-practices description Use when reading or writing Python files py pyproject toml requirements txt Python Best Practices Follows type-first functional and error handling patterns from CLAUDE md This skill covers language-specific idioms only Make Illegal States Unrepresentable Use Python's type system to prevent invalid states at type-check time Frozen dataclasses for immutable domain models python from dataclasses import dataclass from datetime import datetime dataclass frozen True class User id str email str name str created_at datetime Frozen dataclasses are immutable no accidental mutation Discriminated unions with Literal python from dataclasses import dataclass from typing import Literal dataclass class Success status Literal success success data str dataclass class Failure status Literal error error error Exception RequestState Success Failure def handle_state state RequestState None match state case Success data data render data case Failure error err show_error err NewType for domain primi.1.9kinstalls4Web Fetchweb-fetch is a bun-based fetch script in 0xbigboss/claude-code that retrieves a URL, parses HTML with linkedom, selects content-rich elements such as article tags, and converts the result to markdown via TurndownService. Usage is `bun fetch.ts <url>` with a ClaudeCode/1.0 user agent header. Developers reach for web-fetch when agents need live documentation, blog posts, or reference pages as markdown instead of raw HTML. The pipeline covers fetch, DOM parse, content element selection, and markdown conversion for agent-readable output.946installs5Zig Best Practiceszig-best-practices is an agent skill in 0xbigboss/claude-code that steers Claude toward idiomatic Zig when writing systems utilities, native libraries, and performance-sensitive backend services. It highlights explicit allocator usage, comptime features, error union handling, and minimal runtime overhead instead of patterns borrowed from garbage-collected languages. Developers reach for zig-best-practices when scaffolding CLI tools, embedding native extensions, or implementing services that need deterministic memory behavior and zero-cost abstractions. The skill fits code generation and review passes where agents might otherwise introduce hidden allocations, unnecessary copies, or non-idiomatic comptime misses. Use it during Zig module authoring, native library bindings, and low-level API implementations that must compile cleanly with Zig's safety and performance expectations. It encourages choosing general-purpose allocators deliberately, surfacing errors explicitly, and leveraging comptime validation rather than runtime reflection. Teams adopt it when Claude-assisted Zig must match production systems conventions rather than read like translated JavaScript or Go.473installs6Nix Best Practicesnix-best-practices is a Claude Code skill for working with Nix flakes, overlays, shell.nix, and flake.nix files in polyglot repositories. It documents a standard flake.nix skeleton using nixpkgs and flake-utils, follows patterns to deduplicate nixpkgs inputs, and shows how to apply overlay chains including inline final-prev overrides. Dev shell recipes cover mkShell buildInputs, shellHook environment exports, native library PKG_CONFIG_PATH wiring, and direnv use flake integration with NIXPKGS_ALLOW_UNFREE for unfree packages. A binary overlay template fetches platform-specific releases for x86_64-linux, aarch64-linux, x86_64-darwin, and aarch64-darwin with nix-prefetch-url hash workflows. Common commands include nix flake update, nix flake check, nix develop, nix build, and nix run. Developers reach for nix-best-practices when agents scaffold hermetic CI pipelines or cross-language build tooling where config.allowUnfree in flake.nix fails under nix develop.459installs7Tamagui Best Practicestamagui-best-practices is an agent skill from 0xbigboss/claude-code that encodes Tamagui conventions for building shared React Native and web interfaces. It guides agents through cross-platform screens, design tokens, theme setup, and performant styled components so one Tamagui codebase renders consistently on iOS, Android, and web. Developers reach for tamagui-best-practices when implementing product UI with Tamagui instead of maintaining separate React DOM and React Native style systems. The skill reduces inconsistent theming, ad-hoc styling, and platform-specific UI drift during feature development.340installs8Go Best Practicesgo-best-practices is a Claude Code skill from 0xbigboss/claude-code that activates when reading or writing Go files including .go sources and go.mod modules. It enforces type-first design to make illegal states unrepresentable, such as distinct UserID and OrderID string types that the compiler keeps separate at call sites. The skill documents interface-driven behavior contracts, functional error handling, and patterns drawn from the repository CLAUDE.md so generated Go stays idiomatic rather than generic pseudo-Go. Developers reach for go-best-practices during service refactors, CLI tooling, and library work where compile-time safety and standard Go style matter.333installs9Playwright Best Practicesplaywright-best-practices is a Claude Code skill from 0xbigboss/claude-code focused on durable Playwright end-to-end testing patterns. The skill guides agents to prefer stable locators, reusable fixtures, page object structure, parallel CI execution, trace-based debugging, and waits that resist timing flakes. Developers reach for playwright-best-practices when E2E suites grow brittle, CI runs fail intermittently, or new UI flows need tests that survive DOM churn. It complements frontend feature work by standardizing how browser automation is written inside Claude-assisted repositories.329installs10Axe Ios Simulatoraxe-ios-simulator is a Claude Code skill for Apple's iOS Simulator paired with the AXe single-binary CLI installed via Homebrew. It lists simulator UDIDs, performs taps, typing, scroll gestures, home button actions, screenshots, and accessibility inspection through Apple's Accessibility APIs and HID. Developers use it to automate simulator flows and run axe accessibility checks that surface WCAG issues before App Store submission or mobile Safari releases. Commands such as axe list-simulators, axe tap, axe type, axe gesture scroll-down, and axe screenshot form the core workflow. The skill fits mobile engineers validating native iOS UI and mobile web in Simulator without manual device farming.299installs11Electrobun Best Practiceselectrobun-best-practices is a 0xbigboss/claude-code skill guiding agents through Electrobun desktop development with Bun as the main-process runtime and TypeScript across main and renderer layers. Electrobun targets macOS, Windows, and Linux using system webviews—WKWebView, WebView2, and WebKitGTK—instead of bundling Chromium, producing lean binaries often cited around 14MB versus Electron-scale bundles. The skill covers electrobun.config.ts setup, typed RPC between Bun main and webview renderer, BrowserWindow lifecycle, optional CEF bundling on Linux via bundleCEF, and GitHub Actions matrix builds per platform. Reach for electrobun-best-practices when scaffolding a new Electrobun app, hardening renderer security, or preparing cross-platform release artifacts. It suits TypeScript-first developers who want desktop distribution without Rust or a bundled browser engine, accepting per-platform webview rendering differences.280installs12TiltTeaches Claude to use Tilt for Kubernetes-centric local development: author Tiltfiles, wire microservices, enable live reload, manage resource dependencies, and accelerate integration testing for SaaS, API, and CLI-backed backends.249installs13Orbstack Best PracticesCovers OrbStack best practices for fast macOS container development including Docker workflows, local Kubernetes clusters, networking, volume strategies, and performance tuning so agents dockerize stacks efficiently.246installs14Git Best PracticesTeaches Claude Code disciplined Git workflows—feature branches, conventional commits, readable diffs, and pull-request narratives—so changes are easy to review, bisect, and ship without polluting mainline history or hiding breaking edits.213installs15Canton Network ReposOrients developers across Canton Network repositories: locate official SDKs, validator tooling, sample apps, and integration docs; guide clone, dependency setup, and contribution paths for DLT backend and ledger integrations.201installs16Atlas Best PracticesDocuments MongoDB Atlas conventions for Claude Code: project organization, secure connection strings, indexing strategy, staging versus production isolation, and operational defaults so document-database integrations are reliable, observable, and safe under load.194installs17Git Rebase SyncRuns a git rebase-sync workflow: fetch latest main, rebase the feature branch, resolve conflicts systematically, preserve clean history, and prepare an up-to-date branch for PR review and merge.193installs18Testing Best PracticesGuides Claude Code to write maintainable automated tests—clear arrange-act-assert structure, isolated units, realistic integration boundaries, and CI-ready suites—so features are proven correct, refactors stay safe, and releases fail fast on regressions.185installs19E2eThe e2e skill guides Claude through designing and implementing end-to-end tests that simulate real user flows across frontend, backend, and integrations. It focuses on reliable selectors, fixtures, and assertions so teams catch breakage before shipping.171installs20Op CliUses the 1Password op CLI to read vault items, inject secrets into env files and CI pipelines, reference op:// URIs in configs, and enforce rotation workflows without embedding credentials in source code.161installs21Tiltuptiltup supports quick prototyping of tilt-responsive or motion-rich mobile interfaces so teams can feel interactions early. It helps validate whether parallax, device-orientation, or animated layouts merit full engineering before the build phase.154installs22Spec Best Practicesspec-best-practices defines how to create, review, and update SPEC.md files. A developer uses it when authoring a spec or entering a spec gate, ensuring the file has a problem/solution narrative, a domain model, stable REQ-* requirement IDs, invariants, non-goals, and acceptance criteria. It requires reading code before writing spec content and colocating specs with the code they describe.86installs23Git Worktree Tidygit-worktree-tidy performs routine hygiene on bare-repo plus worktree layouts: it fetches origin, prunes remote-tracking refs, deletes stale local branches and worktrees, and fast-forwards important branches. A developer runs it to clean up after squash- and rebase-merges. It verifies ship status against the forge's PR merge state because a deleted upstream does not prove the work merged, and it requires confirmation before any destructive action.79installs24Zmxzmx is a Claude Code skill that gives an agent conventions for managing long-lived shell processes with the zmx session tool. It covers deriving session names from the git root, starting processes idempotently, reading scrollback for errors or readiness, waiting on completion, and killing only the current project's sessions. A developer uses it when an agent needs to keep a dev server, file watcher, test watcher, or tilt process running across turns without blocking the shell.76installs25Improveimprove runs a structured improvement pass over work done in the current session, surfacing concrete suggestions each grounded in a specific observation such as a file path or an error seen. A developer invokes it after completing a task, before handoff, or when asked what could be improved. It sorts suggestions by impact, marks speculative ones explicitly, and integrates with ralph loops and handoffs.71installs26Ios Device Toolkitios-device-toolkit drives a physical iOS device over USB using pymobiledevice3. A developer uses it to take screenshots, tail syslog, pull crash reports, install and launch apps, transfer sandbox files, monitor performance, and forward TCP ports. It documents iOS 17+ requirements such as the tunneld daemon and the DVT screenshot path, plus common connection gotchas.26installs27Brief Best Practicesbrief-best-practices guides an agent to create, review, or update a BRIEF.md, the document that codifies what 'good' and shippable means for a given surface. A developer uses it when work will loop and the agent needs a bar it can verify its own iterations against without interrupting. It defines seven required concerns (Bar, Dimensions, Floors, Oracle, Never, Decisions, Boundary) and the rule that the verifier must be independent of the maker.13installs