# Getting Started

This section gives the shortest path into Monata. For the full workflow, read
the {doc}`../user-guide/index`.

## Install Monata

Install Monata from PyPI:

```bash
pip install monata
```

Install optional first-party technology metadata and bundled public model-card
assets only when you need them:

```bash
pip install monata-techlib
```

`monata-techlib` is separate from core Monata because it carries redistributed
third-party model data and notices. See {doc}`../user-guide/models-and-technology`
for the package boundary.

For local development:

```bash
git clone https://github.com/lizhangmai/monata.git
cd monata
pip install -e ".[dev]"
```

## Install a Simulation Backend

Monata can organize libraries and generate views without a simulator, but you
need at least one backend to run simulations.

- ngspice: current native subprocess backend and the best default for local
  simulations
- OpenVAF: compiles Verilog-A models into OSDI artifacts used by Monata's
  ngspice workflows
- Xyce: deferred backend for larger or parallel simulation flows

The public Monata package does not bundle simulator binaries. The recommended
path is to use the public
[`lizhangmai/skills`](https://github.com/lizhangmai/skills)
`monata-sim-env` and `conda-build` skills from your coding agent, so the agent
builds or reuses a local conda channel for circuit tools and creates the pixi
environment for you.

In Claude Code, install the skill through the plugin marketplace:

```text
/plugin marketplace add https://github.com/lizhangmai/skills
/plugin install monata-sim-env@lizhangmai
/plugin install conda-build@lizhangmai
```

In Codex or another agent, install both skills with the open skills installer
or that agent's normal skill-install flow:

```bash
npx skills add lizhangmai/skills --skill monata-sim-env --skill conda-build
```

Then ask Codex, Claude Code, or another skill-aware coding agent:

```text
Use the monata-sim-env skill to set up this Monata environment.
CONDA_BUILD_OUTPUT_DIR=<absolute-path-you-choose>
```

Replace `<absolute-path-you-choose>` with a real absolute path before sending
the prompt. If the prompt does not include `CONDA_BUILD_OUTPUT_DIR=...`, the
agent should ask for it before running build, pixi, or install commands. The
skill inspects the Monata workspace before choosing tool packages; the current
Monata baseline is `ngspice` plus `openvaf-r`.

See {doc}`../toolchain/external-tools` for the public installation boundary,
extra circuit packages, and fallback source-build guidance. See
{doc}`../toolchain/ngspice` for backend details.

## Create a Library

```python
from monata import LibraryRegistry

reg = LibraryRegistry()
lib = reg.create_library(
    path="work/cmos_cells",
    name="cmos_cells",
    tech_model_paths=[],
)
cell = lib.create_cell("inverter")
```

## Next Step

Continue with {doc}`../user-guide/concepts` to understand how libraries, cells,
views, models, simulations, measurements, and optimization fit together.