dotfiles

Finalize a merged PR by switching to main, pulling, and removing the merged feature branch locally and remotely

Run full production audit on the current project

Create a GitHub pull request following project conventions. Use whenever a pull request is about to be created, whether the user asked directly or another skill or workflow (e.g. superpowers finishing-a-development-branch) reached its PR step. Always takes precedence over inline gh pr create instructions in other skills. Handles commit analysis, branch management, the repo's PULL_REQUEST_TEMPLATE, and PR creation using the gh CLI tool.

DDD patterns - Entities, Aggregate Roots, value objects, Repositories, Domain Services, Domain Events, Specifications. Use when designing domain layer, creating entities, repositories, or domain services.

Django architecture patterns, REST API design with Pydantic for validation and serialization, ORM best practices, caching, signals, middleware, and production-grade Django apps.

Loop project checks and pre-commit, dispatching a fixer subagent per failure, until everything passes or 5 iterations. Use in the Verify stage or standalone when checks fail.

Interview the user relentlessly about a plan or design until reaching shared understanding, resolving each branch of the decision tree. Use when user wants to stress-test a plan, get grilled on their design, mentions "grill me", or right after superpowers:writing-plans produces an implementation plan.

Investigate a Sentry exception down to root cause and propose a fix. Use whenever the user pastes a Sentry issue URL or short ID (e.g. https://org.sentry.io/issues/123456 or PROJECT-1AB2), mentions a Sentry issue, asks "what is causing this exception", or wants a production error diagnosed. Also use when the user asks to "investigate", "debug", or "root-cause" an error that lives in Sentry, even if they don't say the word Sentry.

Design LLM applications using LangChain 1.x and LangGraph for agents, memory, and tool integration. Use when building LangChain applications, implementing AI agents, or creating complex LLM workflows.

Autonomously deep-scan entire codebase line-by-line, understand architecture and patterns, then systematically transform it to production-grade, corporate-level professional quality with optimizations

Python type safety patterns, generics, protocols, and advanced type annotations. Use when adding type annotations, implementing generic classes, defining structural interfaces, or configuring type checkers.

Review current branch changes for quality and security

Save a high-density summary of the current session to .claude_sessions.md

Start the feature development pipeline

Initialize, ingest, query, and lint a Karpathy-style personal wiki inside an Obsidian vault. Use when the user says "initialize the wiki", "start ingestion", "ask the wiki", "what does the wiki say about...", "do wiki maintenance", or similar.

Use when writing prose humans will read—documentation, commit messages, error messages, explanations, reports, or UI text. Applies Strunk's timeless rules for clearer, stronger, more professional writing.