Coding standards, frameworks, refactoring, and code review (18.4K skills)
Review code changes with GPT Sol in a controlled loop before or after opening a pull request. Track findings by severity, make minimal fixes, rerun checks, and continue only while meaningful issues remain, so the review stays focused on the requested change.
Take a requested change through implementation, proof, review, and pull request readiness, then hand it back without merging. It follows the full-send workflow for testing and review, checks the live pull request state, and reports whether it is ready, awaiting approval, or blocked.
Get parallel feedback on a request from three delegates using models from different providers. The skill gathers their findings, summarizes agreements and disagreements, and suggests next steps; code reviews also include a score from each delegate.
Format comments posted on GitHub on the user’s behalf with clear attribution, then review the full Markdown for private details before posting. The skill covers PR comments, inline reviews, replies, and explanations for resolving feedback.
Prepare and open a concise pull request by checking for an existing PR, reviewing the branch diff, and writing a clear title and problem-focused description. The skill also checks the submission text and attachments for secrets or private details.
Carry a requested code change from implementation through direct proof, review, pull request updates, and merge when the required checks pass. The workflow also tracks review findings and asks the user only when progress is blocked or a decision is outside its permission.
Take responsibility for finishing an existing PR, branch, or agent task. Review its history, current code, feedback, and uncommitted work, then make focused changes, verify realistic workflows, and preserve the original contributors’ credit.
Review an older pull request against the current codebase, decide whether its goal still fits, and recommend modernization, rework, or closure. For useful changes, rebase the branch onto its latest base and report conflicts, checks, and results.
Guides you through setting up the Serena MCP server for a chosen project or user scope. It helps identify your client, check prerequisites, configure the connection, initialize project indexing, and verify that Serena tools are available.
Trace a bug backward through its callers to find where invalid data or behavior first originated. The skill suggests adding logs and stack traces when needed, then fixing the source and adding checks at relevant layers.
Set up Codemap CLI for a project or user account, install it, verify it, and add configuration guidance and hooks. Codemap presents codebase structure, dependencies, and changes to help agents navigate a project.
Trace a problem from its visible symptoms to underlying technical or process causes by asking “why” in repeated steps. The skill helps validate causes by working backward, explore multiple branches, and suggest solutions that address the root cause.
Delegate independent implementation tasks or investigations to separate agents, then review their results at checkpoints. The workflow supports sequential work for related tasks and parallel work for unrelated problems, with verification and integration steps to catch issues.
Prioritize changed repository files for human review by combining agent assessments of the change story, impact, failure risks, and expectations. It produces a focused file list, key facts, a random sample of changed files, and declarative files for review.
Set up Context7 MCP and add relevant documentation library IDs to the appropriate CLAUDE.md file. Choose project-shared, project-local, or user-wide configuration, and identify technologies from the supplied input or project structure when needed.
Analyze a GitHub issue and the related project code, then create a structured technical specification covering the problem, proposed approach, implementation and test plans, affected files, and success criteria.
Compare recent Git changes with the project’s FPF knowledge base to find context drift, evidence tied to changed files, and decisions that may need review. The skill reports findings and updates the baseline commit for the next audit.
Applies an incremental approach to coding, refactoring, system design, and workflow changes. It encourages error prevention through types and early validation, consistency with existing project patterns, and solutions sized to current needs instead of speculative complexity.
Load unresolved review threads from a GitHub pull request and turn them into focused Markdown task files under `.specs/comments/`. Group related feedback without duplicating work, preserve actionable suggestions, and leave GitHub comments and PRs unchanged.
Analyze a project’s tools and settings, suggest suitable Claude Code hooks, then configure and validate them. Hooks are commands triggered by events such as file edits or before tool use; examples include formatting code, running checks, and blocking sensitive edits.
Review staged changes and create concise commits using emoji paired with Conventional Commit types such as feat, fix, and docs. The workflow can run project checks, suggest splitting unrelated changes, and follow a branch naming pattern when starting from the main branch.
Review completed work from requirements, architecture, and code quality perspectives. Specialized judges assess it independently, verify their findings, and compare disagreements before producing an evidence-based report with prioritized issues and recommendations. The review does not change files.
Work on several Git branches in separate directories using Git worktrees. The skill covers creating and cleaning up worktrees, reviewing pull requests without interrupting current work, running isolated tests, and comparing or selectively combining changes.
Turn recurring agent mistakes or behavior gaps into focused `.claude/rules` files. The guidance uses incorrect and correct examples to make expectations clear, and helps distinguish always-on rules from task-specific skills or broad project notes.