> ## Documentation Index
> Fetch the complete documentation index at: https://docs.malmhq.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Quickstart

> Install the public CLI and choose generation or evaluation.

# Quickstart

Install the public CLI, then choose the workflow that matches the work at hand.

## Goal

```text theme={null}
OpenAPI + reviewed tasks → CLI or MCP package and Skills
CLI archive + plan → evidence
```

## 1. Install

```bash theme={null}
npm install --global @malmhq/intermesh-cli
```

Requires Node.js 22 or newer and npm 11 or newer.

## 2. Generate a CLI and Skills

Create a workspace, scaffold a named generation, and review its task file:

```bash theme={null}
inter generate setup --name my-api --openapi ./openapi.json --target cli \
  --model provider/model-name --credential-env MODEL_API_KEY
# Review .intermesh/generations/my-api/generation.json
inter generate run --name my-api
```

Generation checks every reviewed task before it creates and verifies the package. It
requires an explicit provider/model selection and its named credential variable.

## 3. Evaluate an existing CLI

Run `inter eval setup my-evaluation` to install the project Skills. If the
configuration is missing, use the `inter-eval-setup` Skill to create
`.intermesh/evaluations/my-evaluation/eval.json`, then run setup again.

```bash theme={null}
inter eval setup my-evaluation
inter eval run my-evaluation
inter eval results my-evaluation
```

Evaluation requires Docker, its Harbor/Python runtime, the selected model credential, and
the dedicated target credential named by the plan. Review the plan and use
`--yes` only after authorizing that exact run.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.