Aller au contenu
TakoAPI
Retour aux skills

google-gemini-media

par xsir0

Image & Video Generation
3.0kTéléchargements rapportés par clawskills.sh
5Étoiles
1Vues
0 mention J'aimeConnecte-toi pour aimer ce skill
—Pas encore noté
Ta note

Résumé

Use the Gemini API.

Description

Ce que fait ce skill

Consolidates six Gemini API capabilities into Node.js and REST templates: image generation and editing (Nano Banana), video generation (Veo 3.1), text-to-speech, and understanding workflows for images, video, and audio. Covers model selection, inline vs Files API input routing, and binary output handling for each capability.

Combining generation and understanding in a single API family removes the need to integrate separate services for image, video, audio, and TTS tasks.

Quand l'utiliser

  • Generating and iteratively editing product images from text prompts
  • Creating podcast or audiobook narrations with controllable voice style and accent
  • Summarizing long videos with timestamped event extraction
  • Transcribing meeting audio and redelivering it as a polished narration
  • Validating AI-generated images by running them through vision understanding in the same pipeline

Exemple de workflow

ENTRÉE

User asks: Generate a cinematic 8-second video of a cat astronaut on the moon and produce a narration script from it

AGENT
  1. 1

    Call Veo 3.1 with the prompt and resolution config, receiving an async operation ID

  2. 2

    Poll the operation with exponential backoff until generation completes, then download the MP4

  3. 3

    Upload the video via the Files API to obtain a reusable file URI

  4. 4

    Call video understanding with a timestamped summary prompt to extract a storyboard

  5. 5

    Pass the storyboard text to Gemini TTS to generate a narration PCM audio file

SORTIE

An MP4 video file and a PCM audio narration summarizing the video's key scenes with timestamps

Prérequis

  • GEMINI_API_KEY environment variable

Installation

clawhub install xsir0/google-gemini-media

Exécute la commande CLI, ou colle l'instruction dans Claude Code, Cursor, Codex ou l'agent de code que tu utilises.

Liens