banner_3

Pantograph is a local-first headless framework with an optional GUI desktop app that provides a unified local inference API, node based workflows, system resource aware runtime schedualing, real-time observability and diagnostic tracing of Ai pipelines. Functionality is provided via language bindings for Python, C#, Elixir, or the native Rust API surface.

1. Clone the repository.
2. Install dependencies:

   npm install

3. Run the desktop app:

   npm run dev:desktop

• Node.js (for npm)
• Rust toolchain (cargo, rustc)
• Tauri system libraries for your OS

Install Rust:

curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh
source "$HOME/.cargo/env"

Install Tauri system dependencies:

# Debian/Ubuntu
sudo apt install pkg-config libsoup2.4-dev libjavascriptcoregtk-4.0-dev
# Fedora
sudo dnf install pkgconf-pkg-config libsoup-devel javascriptcoregtk4.0-devel
# Arch
sudo pacman -S pkgconf libsoup2 webkit2gtk

Install project dependencies:

npm install

npm run dev:desktop

npm run dev

The Vite dev server binds to 127.0.0.1 by default. For an intentional LAN preview, set PANTOGRAPH_VITE_HOST=0.0.0.0 when starting the server.

npm run build:desktop

./launcher.sh --help
./launcher.sh --test
./launcher.sh --release-smoke

• External OpenAI-compatible server (for example LM Studio)
• Bundled llama.cpp sidecar with local model files

• Node.js + npm matching .node-version and package.json
• Rust toolchain matching rust-toolchain.toml
• Python matching .python-version for Python-backed smoke paths
• Tauri system dependencies (above)

Pinned toolchain ownership and update policy are documented in docs/toolchain-policy.md.

# Lint (configured scope)
npm run lint
# Full lint scan
npm run lint:full
# Critical anti-pattern gate (src/ + packages/)
npm run lint:critical
# Focused Svelte accessibility gate
npm run lint:a11y
# No-new-debt gate for critical anti-patterns and decision traceability
npm run lint:no-new
# Rust formatting audit
npm run format:check
# Type check
npm run typecheck
# Tests
npm test
# Runtime separation guard (no compile-time Python linkage)
npm run test:runtime-separation
# Opt-in BEAM / Rustler host smoke
npm run test:rustler-beam-smoke
# Configured quality gates
npm run check
# Canonical local quality gate
./launcher.sh --test

Testing placement, cross-layer acceptance requirements, and release-smoke CI strategy are documented in docs/testing-and-release-strategy.md.

Python-backed model execution is intentionally out-of-process and externally provisioned. See docs/python-runtime-separation.md for configuration and migration details. For a local diffusion worker smoke path, run:

./.venv/bin/python scripts/diffusion_cli_smoketest.py --model-path /path/to/tiny-sd-turbo

Pantograph exposes a Rust-first headless workflow API for host integrations through crates/pantograph-workflow-service:

• workflow_run
• workflow_get_capabilities
• workflow_get_io
• workflow_preflight
• create_workflow_session
• run_workflow_session
• close_workflow_session
• workflow_get_session_status
• workflow_list_session_queue
• workflow_cancel_session_queue_item
• workflow_reprioritize_session_queue_item
• workflow_set_session_keep_alive

Integration boundary:

• Headless hosts should integrate with the core API/service crate directly.
• src-tauri commands are desktop app transport adapters, not the headless API.
• HTTP binding exports are opt-in frontend adapters for modular standalone GUI hosting (frontend-http in UniFFI and Rustler).
• Recommended headless flow: workflow_get_io -> workflow_preflight -> workflow_run.

1. Create a focused branch for one logical change.
2. Follow coding, tooling, accessibility, and documentation standards.
3. Run ./launcher.sh --test and relevant targeted checks before opening a PR.
4. Use Conventional Commits for all commits.

Workspace crates declare MIT OR Apache-2.0 in Cargo metadata. Review individual package metadata for any exceptions.