Pull request creation often misses branch checks, summaries, and test plans. This skill guides Claude, Codex, and Claude Code through GitHub PR preparation.

Many AI-built interfaces look generic and forgettable. This skill guides Claude, Codex, and Claude Code toward stronger visual direction and production-ready frontend output.

Visual details can be hard to turn into clear written findings. This skill helps Claude, Codex, and Claude Code inspect images and report concise observations.

Lovable projects need GitHub sync plus separate backend deployment steps. This skill provides structured prompts and checks for Supabase functions, migrations, secrets, and project maps.

Teams need a small skill to confirm that marketplace loading and prompt routing work. This skill provides a minimal prompt for quick integration checks.

Large repositories make it hard to find the right files and symbols quickly. This skill guides Claude, Codex, and Claude Code toward concise codebase discovery.

Many coding tasks need careful edits that match an existing project. This skill directs Claude, Codex, or Claude Code to implement, verify, and summarize targeted changes.

Developers need fast answers from reliable documentation without losing context. This skill helps Claude, Codex, and Claude Code find references, explain APIs, and summarize version caveats.

Complex features often fail when teams skip requirements, tradeoffs, and rollout planning. This skill helps Claude, Codex, and Claude Code produce structured architecture plans before implementation starts.

API work often needs concise direction before implementation starts. This skill supplies a minimal API-focused prompt context for Claude, Codex, and Claude Code.

Claude Router users need a quick way to understand routing volume and estimated savings. This skill reads the local router stats file and presents route distribution, exceptions, and savings in a clear report.

Automatic model selection can choose a model that does not match your cost, speed, or depth needs. This skill lets you manually route a prompt to Haiku, Sonnet, or Opus through simple command arguments.

Accumulated learning data can become stale, noisy, or inappropriate for a new project phase. This skill guides Claude, Codex, or Claude Code through a confirmed reset of knowledge files, cache, session context, and state.

Continuous learning can keep collecting insights after a focused session has ended. This skill turns that mode off while keeping existing knowledge and manual learning commands available.

Long AI sessions can lose useful decisions, patterns, and observations. This skill enables a router state flag so compatible systems can extract insights automatically.

Important project lessons often disappear after a coding session ends. This skill extracts reusable insights from the current conversation and records them in local knowledge files.

Project knowledge can become hard to track across sessions and files. This skill gives Claude, Codex, and Claude Code a concise status view of stored learnings.

Creating useful agent skills requires clear triggers, safe structure, and practical examples. This skill guides users from idea to finished skill files.

Code reviews often miss security, performance, testing, and maintainability issues under time pressure. This skill gives Claude, Codex, and Claude Code a structured workflow and checklist for thorough review feedback.

Complex coding tasks can fail when scope, risks, and verification are unclear. This skill helps Claude, Codex, and Claude Code create concise plans before implementation.