Coding standards, frameworks, refactoring, and code review (15.8K skills)
Create or improve Claude skills that package specialized knowledge, repeatable workflows, and tool instructions. The guide covers concise instructions, reusable scripts and resources, template initialization, validation, and packaging into a distributable .skill file.
Diagnose file write failures by checking permissions, disk space, paths, and file locks. Get corrective steps before retrying, or check conditions before important writes to catch problems such as missing folders and read-only files.
Connect external services and databases to Claude Code plugins through MCP, a protocol for exposing tools to agents. The skill guides server configuration, connection choices, authentication, tool permissions, and testing so plugins can use those services.
Keeps existing requests for OpenAI documentation guidance working by directing them to the local openai-docs skill. It reuses the canonical instructions, references, and scripts while preserving compatibility with older callers.
Turn existing logs, measurements, or user reports into a prioritized performance improvement plan. It ranks frontend, backend, database, and infrastructure changes by expected impact and effort, and suggests validation steps for checking whether improvements worked.
Find current official documentation for building with OpenAI products and APIs. Provides source-linked guidance on Codex, APIs, SDKs, and model capabilities or limits, using dedicated documentation tools before searching official OpenAI websites.
Handle review and discussion comments on the open GitHub pull request for the current branch. Collect comments, number and summarize the requested changes, then apply fixes for the comments the user chooses using the GitHub CLI.
Turn Figma screens and components into code that follows an existing project's conventions. The workflow retrieves design details and assets, reuses existing components, and checks the resulting interface against Figma screenshots.
Find the right code for Roo Code evaluations, which test AI coding tasks. This skill explains the execution system, management interface, and public results page, helping agents identify where changes and debugging work belong.
Implement, customize, and debug legends in G2 charts with guidance on rendering, layout calculations, and user interactions. Understand category and continuous legends, adjust their appearance, and investigate sizing or filtering problems.
Organize multi-agent reviews with defined roles for bugs, architecture, integration, and usability. Designed for local Codex workflows with vibe, it combines findings by severity into a shared set of keep, simplify, or remove decisions.
Consult an archived reference for classifying errors, reading stack traces, investigating root causes, and recording reusable fixes. Retained for migration, it documents a debugging process while directing current bug investigation to systematic-debugging.
Review code or pull requests for likely bugs, readability, maintainability, and risks to existing behavior. Supporting scripts and reference checklists help analyze code quality, identify missing tests, and organize findings into a review report.
Turn documentation, code, and specifications into reusable Agent Skills. Define when each skill should activate, organize supporting references, add reproducible examples, and check new or existing skills with templates, a checklist, and a validator.
Design backend systems with clear boundaries between business rules, databases, and external services. Use Clean Architecture, Hexagonal Architecture, and Domain-Driven Design to organize new applications, refactor existing code, and make core logic easier to test.
Investigates existing code before feature work, bug fixes, or refactoring. Finds similar implementations, reusable helpers, naming rules, and data handling patterns so proposed changes follow the project's established conventions.
Find official documentation for a library, framework, or API and match it to the version in use. Get source-backed answers and focused code examples, with any details that could not be verified clearly identified.
Cancel the local compatibility-mode state used by Ralph loops in Codex. The bundled PowerShell script removes active state and reports the current iteration, or reports that no active loop state was found.
Turn development errors and debugging attempts into a structured Chinese report covering causes, fixes, and prevention steps. Create a Git commit that references the report, and update an index for later review.
Compare viable design approaches through different perspectives by routing to vibe’s Dialectic Mode. Receive a concise account of shared conclusions, disagreements and their reasons, overlooked issues, and a recommended combined approach.
Turn feature ideas and behavior changes into agreed designs before implementation. Clarify goals through one question at a time, compare alternative approaches, review the design in sections, and save a design document before moving to implementation planning.
Clean up AI-written code by reviewing branch changes against main. Remove unnecessary comments, excessive defensive checks, and casts that bypass type issues, while preserving legitimate changes and matching the surrounding code style.
Carry a software project from requirements and design through implementation, testing, debugging, and deployment. Tracks progress in files, works through features incrementally, and resumes interrupted development, supporting new projects and broad refactoring efforts.
Turn a coding request into a concise, ordered plan after reviewing project documentation and relevant files. The result defines scope, lists concrete actions, and includes validation, risks, and unresolved questions without changing files.