Viznoir
Render cinema-quality scientific plots and VTK scenes from the agent without opening a desktop viz app.
Overview
io.github.kimimgo/viznoir is a MCP server for the Build phase that renders cinema-quality science visualizations headlessly with VTK and saves images for agent workflows.
What is this MCP server?
- Headless VTK rendering exposed through MCP for agent-driven science visuals
- VIZNOIR_RENDER_BACKEND modes: gpu, cpu, or auto
- Configurable VIZNOIR_OUTPUT_DIR for rendered image files
- PyPI package viznoir at version 0.10.1 with stdio transport
- Project site at kimimgo.github.io/viznoir for usage context
- Package version 0.10.1 on PyPI identifier viznoir
- 3 documented VIZNOIR_RENDER_BACKEND values: gpu, cpu, auto
- Transport type stdio per server.schema.json
Community signal: 13 GitHub stars.
What problem does it solve?
Builders waste cycles exporting plots manually when agents could not invoke professional VTK renders from the terminal.
Who is it for?
Solo researchers and technical founders who need high-fidelity science figures in docs without a separate viz team.
Skip if: Marketing-only social graphics or teams that only need simple browser charts with no VTK stack.
What do I get? / Deliverables
Your agent produces VTK-backed visualization files in your configured output directory through MCP tool calls.
- Rendered cinema-quality science visualization image files
- Agent-repeatable VTK pipeline without manual GUI steps
Recommended MCP Servers
Journey fit
Build is where you produce artifacts—figures, explainer images, and dataset visuals—that ship inside repos, papers, or product docs. Docs fits headless image output you drop into READMEs, reports, and validation writeups rather than live UI widgets.
How it compares
Headless VTK render MCP, not a spreadsheet dashboard or generic screenshot skill.
Common Questions / FAQ
Who is Viznoir for?
AI-coding builders and researchers who want agents to generate VTK-based science visuals for documentation and communication.
When should I use Viznoir?
While building when you need reproducible figures for READMEs, validation reports, or technical content alongside your codebase.
How do I add Viznoir to my agent?
Install the PyPI stdio MCP server Viznoir, set VIZNOIR_OUTPUT_DIR, optionally set VIZNOIR_RENDER_BACKEND to gpu, cpu, or auto, then register it in your MCP host.