Build short-form video from raw footage using ffmpeg and local tools — vertical 9:16 reels (TikTok, Instagram Reels, YouTube Shorts) by default, or a 16:9 landscape cut for YouTube. Produces beat-synced montages and talking-head reels with a voiceover spine and b-roll. Handles mixed-source footage, photos with Ken Burns motion, Hyperframes motion graphics, brand grading, Whisper captions, and platform exports. Use for requests to create, edit, assemble, or export reels and shorts, including "make a reel from these clips", "cut this to the beat", "add b-roll over my voiceover", "add motion graphics with Hyperframes", "make a landscape version for YouTube", or "batch process my videos".

Write talking-head video scripts using the Triangle of Yapping — open on the thought or topic, make it land with a metaphor/analogy/example/story, close on application or next steps. Use this skill whenever the user wants to write, draft, fix, or tighten a script for a piece-to-camera video — including "write me a yapping script", "script a talking-head reel", "help me talk about X on camera", "my videos are boring", "I don't know how to open this video", "turn this idea into a script", or "why does nobody watch my talking-head videos". Produces a beat-marked script and a teleprompter file that reel-builder's talking-head mode consumes directly.

Use before implementing or refactoring software when the task requires designing module boundaries, APIs, layers, abstractions, services, repositories, adapters, or architecture. Helps coding agents reduce total system complexity by creating deep modules, hiding implementation knowledge, avoiding leakage and pass-through APIs, comparing alternative designs, documenting interfaces before coding, and critiquing existing architecture.

Use when writing unit tests, integration tests, or end-to-end tests, creating mocks/stubs/fakes, designing a testing strategy, doing TDD, reviewing test quality, fixing flaky tests, or refactoring test suites. Generates risk-focused test plans, picks appropriate test levels, chooses between mocks/fakes/real dependencies, and applies Arrange-Act-Assert patterns with concrete examples.

Security defaults that belong in every FastAPI application from day one. CORS, rate limiting, security headers, trusted hosts, input validation, HTTPS redirect, and request size limits. Apply these whenever you create or modify any FastAPI app -- do not wait for a security review or explicit request. If you are writing app = FastAPI(), you must also add these.

Every UI component must be accessible from the start — not as an afterthought. Apply ARIA landmarks, form labels, keyboard navigation, focus management, live regions, semantic HTML, heading hierarchy, alt text, and color contrast whenever you build or modify any HTML, React component, Vue component, Svelte component, Angular template, or frontend UI. This applies to ALL frontend work — even when the user does not mention accessibility.

Build real-time multiplayer games with PubNub game state sync

Stress-test the primary review with an additional independent reviewer that generates its own findings, compares reviewer conclusions, and identifies issues the primary reviewer may have missed. Use when performing a second opinion or double-check review on a pull request, for medium or high risk PRs, when authoring was heavily AI-assisted, when primary reviewer confidence is low, when findings conflict, or when you need to verify findings with a cross-model or same-model challenger. Supports same-model and cross-model configurations for fair comparison.

Turn many candidate findings from reviewers and verifiers into a small, decision-useful set. Deduplicates, ranks, and suppresses weak findings to consolidate review results into a prioritized, actionable list with severity ratings and merged confidence scores. Use when you need to merge findings, consolidate feedback, prioritize issues, or summarize review output after review passes are complete and before human handoff. Trigger phrases: "consolidate review results", "merge findings", "deduplicate feedback", "prioritize issues from review", "summarize reviewer output". The evidence threshold is the filter — not an arbitrary cap.

Provide an independent critique of a pull request (PR) using a clean reviewer context, identifying bugs, security issues, code quality problems, API misuse, and missing test coverage. Use when performing a code review or pull request review after an evidence pack has been built, for green or yellow risk lane PRs, or as part of a full pipeline for red risk lane PRs. Produces candidate findings (covering correctness, security, and architectural concerns) for downstream synthesis — not final verdicts. Operates as a critic, not a co-author. Common triggers: "review this PR", "code review feedback", "fresh review", "independent review".

NestJS (Nest.js) production patterns for modules, controllers, providers, guards, interceptors, pipes, middleware, JWT, ValidationPipe, microservices, GraphQL, Bull queues, Prisma, and TypeORM. Triggers: NestJS, Nest.js, Nest module, dependency injection, class-validator DTO, exception filter, testing module, GraphQL resolver, Bull queue, microservice client. Uses: Read, Grep, Glob, Bash, WebSearch. Outputs: tier-ordered review checklists and/or concrete code edits with cited rule filenames. Do NOT use for: non-Nest backends (Express/Fastify only with no Nest integration), frontend-only frameworks, generating AGENTS.md, or toolchain setup unrelated to Nest.

Fake clocks / freeze time in tests across every mainstream runtime: freezegun (Python), Jest fake timers + Sinon @sinonjs/fake-timers (JS/TS), timecop (Ruby), java.time.Clock / InstantSource injection (JVM), .NET TimeProvider / FakeTimeProvider, and libfaketime (LD_PRELOAD for any native binary). Covers the language-agnostic discipline - inject or patch the clock, freeze vs tick vs advance vs set-system-time semantics, teardown so fake clocks never leak between tests - plus the shared anti-pattern table (real sleep under a frozen clock, leaked clock state, timezone-dependent assertions). Per-library setup, API, and CI recipes live in references/{python,js,ruby,jvm,dotnet,libfaketime}.md. Use when tests need deterministic control of now(), timers, or timeouts in any language, or when choosing the right fake-clock tool for a stack.

Configures and runs async Python tests with pytest-asyncio: installs the plugin, selects asyncio_mode (auto vs strict), scopes event loops (function/class/module/session), writes async fixtures with @pytest_asyncio.fixture, mocks coroutines with AsyncMock, and tests FastAPI (httpx.AsyncClient + ASGITransport) and aiohttp (aiohttp_client fixture) applications. Use when a Python project contains async def test_ functions, FastAPI/aiohttp endpoints, or any asyncio-based code that needs pytest integration. Do NOT use for general pytest fixture design, parametrize patterns, or conftest.py structure without an asyncio-specific problem (event-loop scoping, mode config, AsyncMock, ASGI client): use python-unit-tests for those.