kanban-video-orchestrator

Orchestrate a multi-agent Kanban video production pipeline from a video brief.

Updated Apr 9, 2026
One-click install
npx skills add https://github.com/MarbleSodas/Mavis --skill kanban-video-orchestrator-marblesodas
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: kanban-video-orchestrator
Source: https://github.com/MarbleSodas/Mavis/tree/main/optional-skills/creative/kanban-video-orchestrator
Command: npx skills add https://github.com/MarbleSodas/Mavis --skill kanban-video-orchestrator-marblesodas

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires argparse, json, os, re, sys, pathlib, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

It eliminates the chaos of turning a vague “make me a video” request into a structured, multi-step production plan that can be executed reliably by specialist agents.

Core Features & Use Cases

  • Adaptive discovery + brief writing: Gathers only the needed inputs (video type, duration, aspect/platform) and produces a contract-style brief with scenes, audio, and deliverables.
  • Team composition + kanban task graph: Chooses the right role archetypes and tool routing for the requested style (e.g., ASCII, Manim, p5js, ComfyUI, TouchDesigner, Blender, comic/pixel styles).
  • Setup + automated execution handoff: Generates and runs a setup script that creates Hermes profiles, configures workspaces, writes TEAM/brief/taste assets, then starts the director’s Kanban decomposition.
  • Monitoring and intervention guidance: Provides the operational monitoring workflow and playbooks for stuck tasks, failures, and re-renders.

Quick Start

Ask for a video like: “Make a 45-second 9:16 product teaser in a retro ASCII-terminal style with voiceover and a music bed, then monitor the pipeline and re-render any failed scenes.”

Frequently Asked Questions about kanban-video-orchestrator

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate a multi-agent video production pipeline from a brief?

A multi-agent pipeline orchestrates specialist agents by decomposing a video brief into a structured kanban task graph. This Skill routes scene beats to rendering agents like ComfyUI, Blender, or Manim, coordinating the full production workflow.

Can I use a Kanban workflow to manage scene planning across different rendering tools like ComfyUI and Blender?

Yes, Kanban scene planning routes tasks to specific rendering tools like ComfyUI, Blender, and Manim. The pipeline dynamically assigns scene beats to specialist role archetypes configured for the requested visual style and rendering approach.

What is multi-agent orchestration for video generation and how does it work?

Multi-agent orchestration for video generation assigns specialized roles to different agents, coordinating them through a Kanban board. It works by translating a video brief into a task graph, routing scenes to appropriate renderers, and monitoring execution.

How do I set up a Hermes workspace for a coordinated video production task?

Setting up a Hermes workspace involves running a generated setup script that creates agent profiles, configures the workspace, and writes brief and team assets. This initiates the director's Kanban decomposition to start the pipeline.

Does this pipeline support retro ASCII-terminal styles and other alternative video formats?

Yes, the pipeline supports alternative video formats including retro ASCII-terminal, pixel art, and comic styles. It adapts tool routing to match the requested style, applying specialized rendering approaches for each visual format.

What should I do if a scene fails during the automated video rendering pipeline?

If a scene fails during the video rendering pipeline, the Skill provides operational monitoring playbooks for stuck tasks and failures. It includes guidance for intervention and automated re-renders to recover the production workflow.