Sanzaru
Wire OpenAI Sora, image, and audio APIs into Claude Code or Cursor via an MCP server so agents can generate multimodal assets in one workflow.
Overview
sanzaru is an agent skill for the Build phase that adds an MCP server for OpenAI multimodal APIs—Sora video, images, and audio—for coding agents.
Install
npx skills add https://github.com/TJC-LP/sanzaru --skill pluginWhat is this skill?
- MCP server exposing OpenAI multimodal APIs to coding agents
- Sora video generation tool surface
- Image generation integration
- Audio generation integration
- Plugin-packaged for agent IDE workflows
Adoption & trust: 6 GitHub stars.
What problem does it solve?
Your agent can write code but cannot call Sora, image, or audio APIs through a standard MCP tool interface.
Who is it for?
Indie builders and small teams building content, demos, or agent products that need programmatic multimodal generation from the IDE.
Skip if: Teams that only need text completions, lack OpenAI multimodal API access, or cannot allow network-backed agent tools.
When should I use this skill?
You need an MCP server for OpenAI multimodal APIs (Sora video, image generation, audio) inside your coding agent environment.
What do I get? / Deliverables
After setup, your agent can request video, image, and audio generations through MCP tools instead of one-off API scripts.
- Configured MCP server reachable from the agent
- Callable tools for video, image, and audio generation
Recommended Skills
Journey fit
Multimodal MCP tooling belongs on Build because you install and configure the server before agents can call video, image, and audio endpoints. Agent-tooling is the right shelf for MCP servers that extend coding agents with new tools—not for shipping SEO or running production monitors.
How it compares
An MCP server plugin for agent hosts—not a standalone creative web app or a Prism workflow-only skill without tools.
Common Questions / FAQ
Who is sanzaru for?
It is for solo and indie developers using Claude Code, Cursor, or similar agents who want OpenAI Sora, image, and audio capabilities exposed as MCP tools.
When should I use sanzaru?
Use it in Build while setting up agent-tooling: install the plugin, configure API access, and validate multimodal tool calls before you embed generated assets in a landing page or demo.
Is sanzaru safe to install?
Review the Security Audits panel on this Prism page, scope OpenAI API keys with least privilege, and understand that the MCP server requires network access to OpenAI services.
SKILL.md
READMESKILL.md - Sanzaru
MCP server for OpenAI multimodal APIs (Sora video, image generation, audio) # sanzaru Source: ./plugin { "name": "sanzaru", "source": "./plugin", "description": "MCP server for OpenAI multimodal APIs (Sora video, image generation, audio)" }