generated: '2026-07-20'
method: searched
source: https://rasa.com/docs/
name: rasa
description: >-
  First-party command-line interface installed with the rasa / rasa-pro Python
  package for scaffolding, training, running and testing Rasa assistants.
install:
  - pip install rasa-pro
  - pip install rasa
commands:
  - name: rasa init
    description: Create a new project with example training data, actions and config.
  - name: rasa train
    description: Train a model using NLU data, stories/flows and the domain.
  - name: rasa shell
    description: Load a trained model and converse with the assistant in the terminal.
  - name: rasa run
    description: Start the Rasa server (HTTP API) to serve a trained model.
  - name: rasa run actions
    description: Start the action server (rasa-sdk) for custom actions.
  - name: rasa interactive
    description: Interactive learning session to build training data conversationally.
  - name: rasa test
    description: Test a trained model on NLU data, stories and end-to-end test cases.
  - name: rasa data
    description: Validate, convert, split, migrate and inspect training data.
  - name: rasa export
    description: Export conversation events from a tracker store to an event broker.
  - name: rasa evaluate
    description: Evaluate model components (e.g. markers) against tracker stores.
  - name: rasa inspect
    description: Launch the inspector UI to debug assistant behaviour (Rasa Pro).
  - name: rasa studio
    description: Sync assistant data between the CLI and Rasa Studio (Rasa Pro).
  - name: rasa license
    description: Manage the Rasa Pro developer/enterprise license key.
key_flows:
  - Scaffold then train: rasa init -> rasa train -> rasa shell
  - Serve API + actions: rasa run --enable-api & rasa run actions