Getting Started
Installation
Set up the Maple CLI, OS daemon, and optional local model backend.
Installation
Maple AI OS is easiest to use from a source checkout today. Build the workspace tools, optionally install a local model backend, and verify the OS with the shipped examples plus CLI surfaces.
Prerequisites
- Rust 1.80 or newer
- Git
- PostgreSQL recommended for persistent PALM storage
- Ollama optional for local model-backed playground use
- Docker optional if you want containerized Postgres or Ollama during development
Install Rust:
curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh
rustup update
Install Ollama if you want a local model backend:
curl -fsSL https://ollama.ai/install.sh | sh
ollama pull llama3.2:3b
Build from source
git clone https://github.com/mapleaiorg/maple.git
cd maple
cargo build --release -p maple-cli -p palm-daemon -p palm
export PATH="$PWD/target/release:$PATH"
maple --help
palmd --help
palm --help
This is the recommended install path for the current repo. It gives you the Maple CLI, OS daemon, operator CLI, and example programs in one checkout.
Current install status
The repository does not document a single published install target for the full product surface yet. The source build above is the accurate path today.
Platform notes
macOS
- Homebrew is the easiest way to install supporting tools.
- Ollama can use Metal-backed local inference on Apple Silicon.
- If you see linker issues after a Rust upgrade, run
xcode-select --install.
Linux
- On Ubuntu or Debian, install standard build tooling before compiling Rust dependencies.
- If you need GPU-backed local models, validate CUDA separately before treating provider issues as Maple OS issues.
Windows
- Use WSL2 for the cleanest developer path today.
- Keep the repo and cargo cache inside the Linux filesystem rather than a mounted Windows path for better build performance.
Verify the install
Start with the binaries:
maple --help
palmd --help
palm --help
Then run a non-daemon example:
cargo run --manifest-path examples/mwl-worldline-lifecycle/Cargo.toml
If you want to verify the daemon-backed path, start PALM in one terminal:
palmd --platform development
And in another terminal:
maple doctor --model llama3.2:3b
If Ollama is not running, maple doctor will tell you that explicitly. The worldline and daemon surfaces are still usable without it.
Troubleshooting
Rust is too old
rustup update
rustc --version
Ollama is installed but not serving
ollama serve
Run that in a separate terminal before retrying PALM playground or maple doctor.
PALM doctor reports PostgreSQL failures
By default, PALM expects:
postgres://postgres:postgres@localhost:5432/maple
Override with PALM_STORAGE_URL, or start the daemon in memory mode through maple daemon start --storage memory.
Build failures after dependency changes
cargo clean
cargo build --release -p maple-cli -p palm-daemon -p palm
maple, palmd, or palm is not found
Call the binaries directly from target/release, or ensure that directory is on your PATH.
Next steps
- Continue to 5-Minute Quickstart
- Author a package in Author Your First Agent Package
- Review the OS shape in Architecture Overview
