AI Media Lab

Videogen command center.

One safe entry point for AI video generation and media automation.

Install status

Open Generative AI, muapi-cli, Seedance SDK, Seedance prompts and Generative Media Skills are staged on hostkey.

Ready

Local browser tools

Clips, thumbnails and stories already run safely.

Staging

Open Generative AI

Build passed and PM2 runs internally.

Needs keys

Seedance / MuAPI / Vadoo

Real generation needs provider credentials.

Project queue
P0 staging

Open Generative AI

Image, video, cinema, lip sync, workflow and agent studio. Installed locally; build/security hardening in progress.

Hostkey skills

Generative Media Skills

Agent skill library staged on hostkey. Script help works; generation safely stops until server-side MuAPI keys are configured.

Hostkey docs

Seedance 2.5 prompts

Camera vocabulary, cinematic examples and prompt templates staged on hostkey as a read-only prompt source.

Hostkey SDK

Seedance 2.5 SDK

Python SDK installed in an isolated hostkey virtualenv. Real generation needs server-side key and quota guard.

Research

Claude Fable 5 use cases

Use cases and benchmarks. Good for examples and templates; not a production dependency.

External SaaS

Vadoo

Auto-publish Shorts/TikToks. Needs account/API review before ecosystem integration.

P1 worker

AI YouTube Shorts Generator

Long-form video to viral Shorts. Best as a background worker connected to Clips.

Hostkey CLI

muapi-cli

CLI runner installed in an isolated hostkey virtualenv. Kept internal because key-management commands can reveal raw keys.

P1 UI

Vibe Workflow

Node-based media workflow builder. Already pulled as Open Generative AI submodule workaround.

P1 script

Text To Video AI

Lightweight script. Good for first backend smoke after credentials are configured.

P2 node pack

muapi-comfyui

ComfyUI nodes for MuAPI. Useful when VISHVA gets a GPU/ComfyUI pipeline.

P2 automation

n8n-nodes-muapi

n8n community nodes. Should integrate only through the existing automation boundary.

P1 design

Open AI Design Agent

Autonomous design agent. Pulled as submodule workaround; needs isolated review before publish.

P2 social

Social Scheduler

Self-hostable social media scheduler. Needs OAuth, queue and posting permission model.

P2 voice

AI Voice Agent

Realtime voice conversations. Needs telephony/audio permission and provider cost controls.

Production shape
vishva.studio/videogen.html ├── browser-safe tools: Clips, thumbnails, stories ├── server job API: /api/videogen/jobs ├── workers: Seedance, MuAPI CLI, Shorts, ComfyUI, n8n ├── secrets: provider keys in server vault only └── outputs: signed storage URLs, never raw provider keys in HTML