Skill UI

Build real-time bidirectional streaming applications using the Gemini Live API. Supports WebSocket-based audio/video/text streaming, voice activity detection, function calling, and session management. Includes SDKs for Python and JavaScript for low-latency AI interactions.

Generative video editing, text-to-video, image-referenced video generation, and first-frame-to-video transition animations using the official google-genai SDK. Supports video editing, style transfer, inpainting/outpainting, and interpolation.

This skill guides the creation and maintenance of a `.geminiignore` file. By systematically filtering out machine-generated noise—such as build artifacts, dependency lock files, logs, and caches—it significantly optimizes the AI agent's context window. This process improves processing speed and, critically, reduces token consumption costs (FinOps).

Generate and edit images and videos using Google's Gemini media models (Nano Banana 2/Pro, Gemini Omni Flash) with cost-approval gates, reference-image support, and detailed logging per call. Supports on-brand generation from reference sets and multi-image fusion for consistent character series.

Audit and optimize websites for platform-specific AI search engines including Google AI Overviews, ChatGPT, Perplexity, Gemini, and Bing Copilot. Provides per-platform checklists, scoring rubrics, and action items to improve visibility and citation rates in generative engine optimization (GEO). Requires network access for site audits; PDF reports need pandoc and headless Chrome.

This skill provides patterns for automating GitHub workflows with AI assistance. It helps in automating PR reviews, setting up issue triage, creating GitHub Actions workflows, and integrating AI into CI/CD pipelines. It also covers automating Git operations like rebases and cherry-picks. Inspired by modern DevOps practices and tools like Gemini CLI, this skill streamlines development processes and reduces manual effort in repository management.

Generate and edit images using Gemini's Nano Banana Pro model (gemini-3-pro-image-preview). Supports text-to-image, image editing, multi-image composition, and Google Search grounding. Provides Python, JavaScript, and REST API examples.

Generate images with Google Gemini's image model straight from a Claude Code session. Create UI placeholders, icons, documentation illustrations, diagrams and concept visuals, saved locally as PNG files with a returned file path.

A token-efficient, high-quality development workflow optimized for Gemini 3.1 Pro High. Covers planning, architecture, TDD, clean production code, technical validation, mandatory visual/UX pixel-perfect checks, and final verification.

This blueprint guides the creation of F.R.I.D.A.Y., a tactical, low-latency voice assistant inspired by Iron Man. It utilizes the specialized Pipecat framework to orchestrate a comprehensive, real-time audio pipeline: STT (OpenAI Whisper/GPT-4o) -> LLM (Gemini) -> TTS (OpenAI Nova). Ideal for developers needing to build advanced, multi-provider, conversational voice applications.

Polis Protocol transforms static AI agent workflows into a self-improving, dynamic team environment. It uses a learning router to assign tasks based on an agent's historical track record and effectiveness on specific domains, rather than fixed roles. It is vendor-agnostic, allowing diverse agents (e.g., Claude, Gemini) to collaborate under a unified, auditable governance structure where agents can propose and vote on amending the protocol's core rules.

PydanticAI is a robust Python framework that integrates Pydantic's type-checking and validation capabilities into Large Language Models (LLMs). It allows developers to build production-ready AI agents that guarantee structured, validated outputs, moving beyond simple raw strings. Key features include supported tool calling, dependency injection for testability, and multi-model compatibility across major providers like OpenAI, Anthropic, and Gemini. Use this when reliability and correct data structure are paramount.