msilvis-ai-skills

Build, run, view, and interact with Android apps on an emulator or connected device. Use when the user asks to build, run, install, screenshot, tap, scroll, swipe, type, inspect, or debug an Android emulator/device.

Collect a batch of bugs / issues / changes from the user, then categorize them and dispatch parallel subagents — one per category — each working in its own git worktree to fix that category's items. Use when the user invokes `/msilvis:bug-bash` or says "bug bash", "let's bug bash", "queue up some fixes", or otherwise signals they want to dump a list of issues and have them worked in parallel.

Diagnose and fix failing CircleCI builds. Use when the user asks to investigate CI failures, fix a broken build, or debug CircleCI pipeline issues.

Full code quality pass — runs /review to find and fix all issues, then runs /simplify with framework-aware focus areas. Use when asked to improve code quality, clean up a repo, or run a full quality pass.

Full code quality pass that auto-merges the PR when CI is green and the change is simple enough. Runs /msilvis:code-quality, then gates a squash-merge on a strict simplicity checklist. Use when the user wants to ship a small PR end-to-end with one command.

Manage Coolify self-hosting platform via Playwright browser automation. Use when the user asks to deploy, check status, manage services, view logs, or configure anything in Coolify.

Free disk space on a Coolify server via the web app's built-in terminal. Use when Coolify warns about low disk space, or the user says "clean up coolify disk", "free space on the server", "coolify is running out of room", or similar. Related to the `coolify` skill — same login, but drives the in-app terminal instead of the dashboard pages.

Help resolve GitHub pull request review comments using the GitHub CLI (`gh`). Use when the user asks to address, resolve, or summarize PR comments, update code accordingly, and optionally draft or post replies back to GitHub.

Implement the GitHub issues created or refined by an audit using multiple focused subagents — pulls the issues, groups them by priority (P0 correctness/security → P3 developer experience) and system area, dispatches non-overlapping workstream subagents, and ships small reviewable PRs with validation. Use when the user says "Implement Audit Issues Using Fable 5 Extra Subagents", "implement the audit issues", "work through the audit issues", or invokes `/msilvis:implement-audit-issues`.

Build, run, view, and interact with iOS apps on the Apple Simulator. Use when the user asks to build, run, install, stream, screenshot, tap, scroll, swipe, rotate, grant permissions, or inspect an app on an iOS/iPadOS/watchOS Simulator.

Create, edit, or manage Claude/Codex/Cursor AI skills. Use when the user wants to add, update, or remove skills or agent instructions.

On-demand performance audit for a Next.js + PostHog app. Gathers slow routes from bundle sizes, Web Vitals RUM, and backend events/errors, then fans out parallel subagents — one per finding, each in its own worktree — that open PRs against the worst offenders. Repo-agnostic; reads `.claude/perf.json` from the current repo for thresholds and signal sources. Stops at "ready for review"; never auto-merges. Use when the user says "run a perf audit", "audit performance", "find slow routes", or invokes `/msilvis:perf-audit`.

Run the test plan checklist of a GitHub PR automatically. Parses the `## Test plan` markdown checklist, executes each item via Playwright (browser flows) or Bash (CLI checks), attempts to fix failures by editing source and re-running, flips passing items to `- [x]` in the PR body, commits + pushes any fixes, and posts a summary comment with pass/fixed/fail/skip results. Use when the user says "run the test plan", "verify this PR", "check off the test cases", or invokes `/msilvis:pr-test-runner` explicitly.

Run a full visual and functional QA pass on a web app. Audits every page on desktop and mobile, dispatches fixes to other agents, re-verifies, and opens a PR when everything passes. Works across any web project — detects routes, dev URL, and validate command from the repo. Supports focused modes (e.g. `silly-classic`) that scope the run to a single feature and run domain-specific functional flows (team creation, scoring, winners) on top of the visual audit.

Implement CSS View Transitions API in Next.js App Router using next-view-transitions. Use when building page transitions, route animations, or the user mentions view transitions in a Next.js project.

Set up a new or cloned repository for local development. Use when the user asks to scaffold a new project, bootstrap a repo, set up local dev environment, configure local domains/proxy, or get a project running locally.

Modernize an existing public website into a new static Next.js site from a user-provided template or starter. Crawls the source URL, downloads assets, extracts content, scaffolds the project, and coordinates parallel subagents for generalized modernization tasks. Use when the user says "modernize this site", "rebuild this site", "refresh <url>", "clone <url> as a new site", or provides a website URL and asks for a rewrite.

Create a draft PR, get user approval on the body, mark it ready for review, and notify Slack. Use when submitting, publishing, or sending a PR for review, or when the user says /msilvis:submit-pr, "send PR", "open PR for review", or "submit for review".