Skill Claude CodeCodex
Forces the AI to detect, confirm, and remember the host operating system before any command execution, file operation, or path construction. Prevents Linux-isms on Windows, wrong path separators, incorrect shebangs, and incompatible shell syntax.
Skill Claude CodeCodex
Measure-first approach — profile before optimizing, Core Web Vitals, bundle analysis, database query optimization. Prevents premature optimization while catching real performance issues.
Skill Claude CodeCodex
Use when the task requires automating a real browser from the terminal (navigation, form filling, snapshots, screenshots, data extraction, UI-flow debugging) via playwright-cli invoked through npx in the project directory.
Skill Claude CodeCodex
Keeps all dependencies, SDKs, virtual environments, and tooling inside the project directory. Prevents polluting the OS drive (especially C: with <30% free) by using project-local installs. Always checks disk space and documents the setup.
Skill Claude CodeCodex
Create a timestamped project backup and inspect the repo's current status before making changes. Use when starting work in any project, before risky edits or refactors, when the user asks to back up or safeguard a codebase, or when continuity matters and code tools should read TODO/status docs before acting.
Skill Claude CodeCodex
Discovers, catalogs, and manages project scripts (install, download, build, setup, clean). Scans conventional script locations, detects AI-generated and user-created scripts, and generates missing standard scripts. Use when setting up a new project, looking for available scripts, or ensuring script consistency.
Skill Claude CodeCodex
Teaches how to receive review feedback properly — verify-before-implement on every comment, different skepticism levels by source, evidence-based pushback, no performative agreement. Prevents blindly implementing review suggestions without understanding them.
Skill Claude CodeCodex
Standardized release workflow — version bump, Keep a Changelog formatting, and CI/CD pipeline integration with auto-tagging and release automation. Works for any language (Python, Rust, Node, Go, etc.). Use when the user asks to cut a release, bump version, update changelog, or prepare release notes.
Skill Claude CodeCodex
Runs a structured Q&A session before starting any ambiguous task. Asks one question at a time using multiple choice first, surfaces assumptions, confirms understanding. Use when the user gives vague requests, incomplete specs, conflicting goals, missing context, or says anything like "build this", "make it work", "fix…
Skill Claude CodeCodex
Ensures code modifications follow safe practices - never remove imports/items without verifying usage, check if truly obsolete, and confirm usage by other modules before removal.
Skill Claude CodeCodex
Use when the user explicitly asks for a desktop or system screenshot (full screen, specific app or window, or a pixel region), or when tool-specific capture capabilities are unavailable and an OS-level capture is needed.
Skill Claude CodeCodex
Perform language and framework specific security best-practice reviews and suggest improvements. Trigger only when the user explicitly requests security best practices guidance, a security review/report, or secure-by-default coding help. Trigger only for supported languages (python, javascript/typescript, go). Do not…
Skill Claude CodeCodex
Analyze git repositories to build a security ownership topology (people-to-file), compute bus factor and sensitive-code ownership, and export CSV/JSON for graph databases and visualization. Trigger only when the user explicitly wants a security-oriented ownership or bus-factor analysis grounded in git history (for…
Skill Claude CodeCodex
Repository-grounded threat modeling that enumerates trust boundaries, assets, attacker capabilities, abuse paths, and mitigations, and writes a concise Markdown threat model. Trigger only when the user explicitly asks to threat model a codebase or path, enumerate threats/abuse paths, or perform AppSec threat modeling.…
Skill Claude CodeCodex
After any batch of changes (edits, skill updates, cross-references, docs), runs systematic validation to catch inconsistencies, missing references, broken links, and syntax errors before declaring done. Prevents the need for follow-up corrections.
Skill Claude CodeCodex
Pre-launch checklists, staged rollouts (canary -> 10% -> 50% -> 100%), rollback plans, feature flag verification. Complements release-changelog (version bump + changelog) by covering the actual deployment and launch process.
Skill Claude CodeCodex
When a skill is loaded, selectively loads the most relevant direct cross-references with strict caps. Preserves core operating discipline without letting cross-references cascade into context bloat.
Skill Claude CodeCodex
Forces writing a specification before any code changes. Defines goals, inputs/outputs, constraints, edge cases, and acceptance criteria before implementation. Use when starting any non-trivial feature, refactor, or bug fix. Prevents solving the wrong problem.
Skill Claude CodeCodex
Enforces RED-GREEN-REFACTOR cycle — write a failing test first, write minimal code to pass, then refactor. Use when implementing new features, fixing bugs, or adding regression coverage. Prevents untestable code and false confidence from untested changes.
Skill Claude CodeCodex
Create or refresh a project TODO checklist using Markdown checkboxes and keep it current as work progresses. Use when a repo has no TODO tracker, when the user asks for a backlog, roadmap, status checklist, or progress board, or when docs exist but do not yet include a canonical checkbox-based task list.
Skill Claude CodeCodex
Detects available build toolchains (MSYS2, Zig, GCC, Clang, Visual Studio) and falls back to a working alternative when none are found. Any implementation scripts follow the project's scripts/ folder convention.
Skill Claude CodeCodex
Run language-appropriate formatter and lint-fix commands for changed files. Use when the user asks to format, lint, auto-fix, or clean code style across Python, JavaScript/TypeScript, Markdown, YAML/JSON, PowerShell, shell scripts, and similar files.
Skill Claude CodeCodex
Diagnose and resolve newly added but unused imports by inferring intent from surrounding code and implementing the missing use case. Use when linting or review reports an unused import and the import appears intentional or recent, and the goal is to complete behavior rather than remove code.
Skill Claude CodeCodex
Use when installing Python tooling, managing Python environments, running Python scripts, or managing Python dependencies. Use command uv for all Python operations instead of pip or python -m venv.