Roadmap

Get Involved · In Progress · Planned Features · Changelog

Get Involved

I maintain the core tools and roadmap, but there's room to collaborate:

• Bug reports — especially reproducible LLM tool misuse
• Small model optimization — help make Ollama/LM Studio work better
• Documentation — improvements and tutorials
• Evaluations — automated end-to-end behavior comparisons of different LLMs and small model mode vs normal mode (basic framework is in place, needs a lot more evals, and manual verification of eval effectiveness and eval results)
• Voice interaction — experimental bidirectional voice chat

Open a GitHub Discussion or reach out directly.

In Progress

1.4.x - More Transforms

• Apply curves
• More flexible randomization: arbitrary min and max, randomly choose between a set of choices (e.g. the notes of a chord)
• More note properties available as variables (e.g. note.index within the MIDI clip, clip.position)
• Ability to apply LFOs relative to the arrangement start instead of clip start
• Code execution feature for the AI to generate custom algorithms for transforms and MIDI note generation

Also planned: Optimizations for better (local) small language model support, and longer conversations in general.

Changelog

1.4 - MIDI Transforms

Enhance the ppal-create-clip and ppal-update-clip tools with flexible MIDI transform features:

• Apply ramps and LFO shapes to velocity, start time, duration, pitch, audio gain and other note and audio properties
• Randomize velocity and timing (and any of the other properties)
• Apply timing patterns like swing to MIDI
• Combine multiple transforms with mathematical expressions
• Changes are applied directly to MIDI clips (the notes are modified). The results can be easily seen and edited by hand.

Also:

• Split arrangement clips at the given position(s)
• Support running Ollama on another computer (Ollama API URL is no longer hard-coded to localhost)

1.3 - Device Control (December 2025 - January 2026)

• Full device control: add/delete/move native devices on any track, read/write parameters, insert into rack chains
• Rack macro and variation management
• A/B comparison for device parameters

Also added support for:

• Arrangement locators
• MIDI clip quantization

1.2 - Audio clip, mixer, and improved Arrangement support (November 2025)

• Audio clip support with a read-samples tool to scan folders for samples
• Track mixer control: gain, panning, and sends
• Arrangement clip positioning and length control

1.1 - Built-in chat UI (November 2025)

Direct API integrations for Google Gemini, OpenAI, Mistral, OpenRouter, LM Studio, Ollama, and more.

1.0 - Support for more LLMs (October 2025)

Expanded features and support for multiple AI platforms.

0.9 - Public Beta with Claude Desktop (July 2025)

Initial public release with Claude Desktop support and a focus on MIDI clip manipulation and basic Live Set management.

Planned Features

1.5 - Customization and Small Model Optimization

• Custom skills, system prompts, tool descriptions, and task-focused personas for experimentation and local model optimization
• Make different personas to focus on different types of tasks, for better results and less context window usage
• Evaluation tools for comparing adjustments to customization settings

1.6 - Harmony

• Chord notation
• Pitches as scale/chord degrees of the current chord (instead of absolute pitches)
• Microtonal support

1.x - Miscellaneous Features

• Take lane support
• Groove support
• Support new Live API features as they become available
• Persist conversations from the built-in chat UI and continue them later
• Voice interface: Speak to Producer Pal instead of typing