Coding standards, frameworks, refactoring, and code review (16.6K skills)
Assess an existing Convex app and turn the findings into a prioritized improvement plan. With explicit approval, update outdated components, address production error monitoring gaps, and reassess the app to show how its score changed.
Turn a one-sentence app idea into a minimal Next.js and Convex project. The skill installs dependencies, starts backend and web development servers, and opens the local development URL for an initial working template.
Guide changes and reviews in datafusion-ffi, which connects separately compiled libraries. Check binary compatibility, memory ownership, and method coverage, and choose unit and cross-library tests for new wrappers, trait extensions, or codec changes.
Review Apache DataFusion pull requests using the project's review guide. Gather the proposed changes, discussion, and automated check results, or apply the same review criteria to your own changes before submitting a pull request.
Creates a separate Git working directory for a task and moves the agent session into it using Worktrunk. Supports choosing another repository, naming the branch, and carrying uncommitted changes into the new workspace.
Manage Git worktrees, separate working folders for branches, with Worktrunk’s wt CLI. Get guidance on command targets, configuration, automated hooks, aliases, commit message generation, and troubleshooting when commands behave unexpectedly.
Guides edits to Worktrunk’s command-line messages, help text, and progress displays. Defines wording, colors, symbols, and where output goes, so command results and status messages follow consistent conventions in terminals, pipes, and redirected logs.
Improve an existing AI agent setup using execution records and user feedback. Identifies recurring problems, turns individual comments into reusable guidance, updates skills and coordination rules, and checks changes while preserving their history.
Build a project-specific team of AI agents with defined roles, reusable skills, and coordinated task flows. Creates configuration files, checks existing setups for overlap, and validates how agents exchange results and handle failures.
Investigate Bifrost GitHub issues by examining reports, related code, and library documentation. Produce an implementation plan with impact analysis and regression tests, including a demonstrated failing test for bugs before requesting approval to fix them.
Prepare Bifrost releases by turning Git history into release summaries and module changelogs. Coordinate version updates across dependent modules, document database migration risks, and include closed issues and outside contributor credits.
Work through unresolved review comments across dependent PRs managed with Graphite. Review branches from the base upward, propose fixes or replies for approval, test changes, and commit each branch locally while preserving stack dependencies.
Add a model pricing field throughout Bifrost, from datasheet parsing and database storage to cost calculations, overrides, documentation, and UI controls. Verify calculation and repeat-sync behavior with tests, then check the repository for missed integration points.
Distribute edits across the appropriate branches in a Graphite stack, a chain of dependent branches. Use manual grouping when automatic absorption misplaces changes, then update commits, resolve restack conflicts, and verify the affected branches locally.
Split a large commit or branch into a Graphite PR stack, a series of dependent review requests. Group changes by purpose, preserve changes that must stay together, and verify that each lower layer builds and passes tests independently.
Review a pull request or code diff from several angles, then have separate agents test each potential finding. Receive a ranked findings table, confidence assessments, a merge recommendation, and concrete follow-up work.
Help a spawned worker check which date-range helpers and external integrations are available and how to call them. The catalog explains environment settings, local broker access, and availability checks so the worker can choose tools for its task.
Move an existing mod to a newer game version or another mod loader while preserving its behavior and values. Organize source comparisons, scripted data extraction, tests, visual checks, and handoff documents for a long migration.
Inspect game code, data files, live memory, and graphics captures to understand the internals needed for a mod. Choose tools for the game’s technology, document findings, and validate unknown file formats by decoding and re-encoding them.
Investigate an installed game before making a mod. Identify its engine, available mod loaders, save locations, anti-cheat protection, and compatibility issues, then record a practical approach and unresolved questions in MODDING_PLAN.md.
Build mods that combine content or gameplay from different games. Choose between content ports, two games exchanging data, embedded game engines, or rewritten rules, with clear integration plans and checks against the original behavior.
Prepare a finished game mod for release by checking packaged files, arranging the required folder structure, writing installation instructions and credits, and drafting a video-led announcement. Includes versioning and platform-specific packaging guidance for user-controlled publishing.
Build modifications for PC games you own, from new weapons and enemies to units and combined game experiences. Covers choosing an approach, creating assets, testing inside the actual game, and recording a short showcase video.
Develop and troubleshoot Apache Beam's YAML SDK. Understand its source files, run data processing pipelines locally or on Dataflow, test changes, and investigate provider issues or mismatched record fields.