Feature Request: Open API / MCP Support for AI-Assisted Quad Cortex & DAW Automation

I would love to see Neural DSP consider opening a deeper external control interface for Quad Cortex — ideally a public API, SDK, or even MCP-compatible interface.

My dream workflow would be:

  1. Import a backing track into Studio One.

  2. Let an AI agent such as Codex analyze the song structure.

  3. Automatically identify sections such as Intro, Verse, Chorus, and Solo.

  4. Create corresponding Scenes on Quad Cortex.

  5. Adjust amps, effects, gain, EQ, delay, reverb, and levels for each Scene.

  6. Automatically generate MIDI automation in Studio One to switch Quad Cortex Scenes during playback.

After that, I could simply say:

“Make the chorus bigger, make the solo less harsh, and add a little more delay.”

Instead of only suggesting parameter values, the AI could directly edit the current Quad Cortex preset.

I think this could be extremely useful for live musicians.

Neural DSP would not necessarily need to build its own AI system. If Quad Cortex exposed deeper control of the signal grid, device parameters, Scenes, presets, I/O, and MIDI through an API or SDK, third-party developers and AI platforms could build these workflows themselves.

CorOS is already becoming much more modular with Device Presets, I/O Presets, Global EQ Presets, Scenes, and MIDI control. Opening those functions to external software could make Quad Cortex one of the most powerful AI-ready guitar platforms available.

Features I would especially love to see:

  • Public Quad Cortex API or SDK

  • Bidirectional external control

  • Access to individual device parameters

  • Access to the signal grid

  • Scene creation and editing

  • Preset loading and saving

  • Deeper MIDI automation

  • Integration possibilities for AI agents such as Codex or ChatGPT

This is probably the future Quad Cortex feature I would be most excited about.

Would anyone else find this kind of workflow useful?

I’ve implemented Quad Cortex MCP here: GitHub - lexasoft123/qc-mcp: Quad Cortex MCP for Claude/Codex · GitHub

It can compose presets, research artist rigs.

Need more reports to make it better

This is almost exactly what I was hoping for!

My ideal workflow is: import a backing track into Studio One → let Codex analyze the song structure → automatically create Intro / Verse / Chorus / Solo scenes on the Quad Cortex → then generate the MIDI scene changes on the Studio One timeline.

I’d also love to be able to tell Codex things like “make the chorus bigger” or “the solo is too harsh” and let it directly modify the QC preset.

I’d be very interested in testing qc-mcp and reporting issues. This is an incredibly exciting project!

I’d love to help test this project, but unfortunately I’m currently using Windows, and I noticed it only supports macOS for now. If Windows support becomes available in the future, I’d be very happy to help test it and provide feedback. Thanks for your great work!

This sounds like a great project. I use MCP connections to Devonthink and Craft every day to analyze documents and notes. Sometimes with Claude Cowork or with LM Studio / Bionic with lokal LLMs. I am on a Mac, so going to try and install the MCP server for QC on the weekend. Does it work with the new CoreOS 4.1.0?

1 Like

I am working on adapting to 4.1.0 now, will release today

I will think about Windows port

2 Likes

4.1.0 support is released

1 Like

Thank you so much for considering it!
Imagine if an AI could actually “listen” to audio and evaluate tones—it would be a game-changer for guitarists. Really appreciate all your hard work on this exciting project!

Hi all
Done: Quad Cortex MCP

Both windows and macos, easy app to start