What problem does it solve? Running AI models like FLUX, Veo, or Seedance normally requires managing separate provider APIs, GPU infrastructure, and authentication for each service. This Skill unifies access to 150+ cloud AI apps through a single CLI, so you can generate images, video, audio, and run AI search without any GPU or per-provider setup. ## Core Features & Use Cases - App Discovery: Search the inference.sh catalog with infsh app list --search to find the correct app ID before running anything. - Media Generation: Run image models (FLUX, Gemini, Grok, Seedream), video models (Veo, Seedance, Wan), TTS, and avatar/lipsync apps with a single JSON input command. - Local File Uploads: Pass local file paths directly in the input JSON and the CLI uploads them automatically for tasks like upscaling, image-to-video, and avatar generation. - Use Case: A user asks for a short drone-style video clip. The agent searches for video apps, runs google/veo-3-1-fast with a JSON prompt, and returns the generated video URL for inline display. ## Quick Start Ask the agent to generate an image of a sunset over mountains using inference.sh, and it will search for the right app and run it with structured JSON output.
