video-ad-specs

Generate AI images with 50+ models (FLUX, Gemini 3 Pro Image, Grok Imagine, Seedream, Reve) via the inference.sh CLI, covering text-to-image, inpainting, LoRA, image editing, and upscaling.

Generate AI videos (text-to-video, image-to-video, lipsync, upscaling) with 40+ models like Google Veo, Seedance, Wan, and Grok via the inference.sh CLI. Covers social videos, marketing clips, explainers, and AI avatars.

Umbrella skill for running 250+ AI apps through the inference.sh CLI — image and video generation, LLM calls, web search, 3D, and Twitter automation — from a single no-GPU-required install.

Install and authenticate the inference.sh (`infsh`) CLI so that agents can run any of 250+ cloud AI apps — image, video, LLM, search, 3D, and more — without needing a local GPU.

Run web search and content extraction through Tavily (Search Assistant, Extract) and Exa (Search, Answer, Extract) via the inference.sh CLI for research, RAG, and fact-checking workflows.

Execute arbitrary Python 3.10 code in a sandboxed inference.sh environment pre-loaded with NumPy, Pandas, Matplotlib, Playwright, MoviePy, Pillow, OpenCV, trimesh, and 100+ other libraries.

Render MP4 video from React/Remotion TSX component code via the inference.sh CLI with configurable resolution, FPS, duration, and codec — supports `useCurrentFrame`, `spring`, `interpolate`, `Sequence`, etc.

Automate Twitter/X through the inference.sh CLI — post tweets with media, like, retweet, delete, send DMs, follow users, and fetch posts/profiles via ready-made `x/*` apps.

Playwright-backed browser automation for AI agents via inference.sh, using an `@e` ref system to open pages, click, fill forms, screenshot, record video, and re-snapshot between navigations.

Design high-converting SaaS/product landing pages using an above-the-fold formula, hero section layout, CTA psychology, social proof placement, mobile rules, and F-pattern reading.