Interactive Prompts

Q: How do I skip prompts in CI?

if os.environ.get("CI"): use defaults or --yes / --non-interactive flag.

Q: How do I handle Ctrl+C?

questionary returns None on cancel. Check and exit gracefully.

Q: Password input?

questionary.password("API key:").ask().

Q: Autocomplete?

questionary.autocomplete("Path:", choices=file_list).ask().

Q: Testing interactive CLIs?

Mock questionary.text(...).ask() return values in pytest.

Q: Default selections?

questionary.select(..., default="FastAPI").ask().

Interactive prompts collect user input for wizards, confirmations, and guided setup. Use Click's built-in prompts for simple cases; questionary for rich selection UIs.

Recipe

import questionary
 
name = questionary.text("Project name:").ask()
framework = questionary.select("Framework?", choices=["FastAPI", "Django", "Flask"]).ask()
confirmed = questionary.confirm("Proceed?").ask()

When to reach for this:

Project scaffolding wizards (cookiecutter-style)
Destructive operations needing confirmation
Multi-step setup where flags would be unwieldy

Working Example

import questionary
from rich.console import Console
 
console = Console()
 
def run_setup_wizard() -> dict:
    if not questionary.confirm("Run interactive setup?").ask():
        raise SystemExit("Aborted")
 
    answers = {
        "name": questionary.text("Project name:", validate=lambda t: len(t) > 0 or "Required").ask(),
        "framework": questionary.select(
            "Web framework:",
            choices=["FastAPI", "Django", "Flask"],
        ).ask(),
        "database": questionary.checkbox(
            "Features:",
            choices=["postgres", "redis", "celery"],
        ).ask(),
        "deploy": questionary.select(
            "Deployment target:",
            choices=["Docker", "Lambda", "VPS"],
        ).ask(),
    }
 
    if not questionary.confirm(f"Create {answers['name']} with {answers['framework']}?").ask():
        raise SystemExit("Aborted")
 
    return answers
 
if __name__ == "__main__":
    config = run_setup_wizard()
    console.print(f"[green]Creating project: {config}[/green]")

What this demonstrates:

Text, select, checkbox, and confirm prompt types
Validation on text input
Abort on cancellation (Ctrl+C or No)
Rich output after wizard completes

Deep Dive

Prompt Selection

Library	Best for
`click.confirm/prompt`	Simple confirm/text
`questionary`	Select, checkbox, autocomplete
`prompt_toolkit`	Full custom REPL
`InquirerPy`	Alternative to questionary

Gotchas

Interactive prompts in CI - hangs forever. Fix: detect CI=true or --yes flag to skip prompts.
No default for destructive confirm - accidental yes. Fix: default=False on confirm().
Password echo - visible on screen. Fix: questionary.password() or click.prompt(hide_input=True).
Prompts without non-interactive fallback - breaks scripting. Fix: every prompt needs a CLI flag equivalent.
Unicode on Windows terminals - display issues. Fix: Rich Console or enable UTF-8 terminal.

Alternatives

Alternative	Use When	Don't Use When
CLI flags only	Scriptable automation	Guided first-time setup
web-based setup	Complex configuration	Terminal-only tool
config file	Repeatable deploys	Initial onboarding

FAQs

questionary or click.prompt?

click.prompt for simple text/confirm. questionary for select, checkbox, autocomplete.

How do I skip prompts in CI?

if os.environ.get("CI"): use defaults or --yes / --non-interactive flag.

How do I handle Ctrl+C?

questionary returns None on cancel. Check and exit gracefully.

Password input?

questionary.password("API key:").ask().

Autocomplete?

questionary.autocomplete("Path:", choices=file_list).ask().

Can I style prompts?

questionary supports style dictionaries for colors.

Testing interactive CLIs?

Mock questionary.text(...).ask() return values in pytest.

Multi-page wizards?

Chain prompts in functions; show progress with Rich.

prompt_toolkit when?

Building a full REPL or syntax-highlighted input. Heavier than questionary.

Default selections?

questionary.select(..., default="FastAPI").ask().

Click - built-in confirm/prompt
Rich Output - formatted wizard output
Config & Environment - post-wizard config
CLI Best Practices - interactive vs scriptable

Stack versions: This page was written for Python 3.14.0, FastAPI 0.115+, Django 5.2, Flask 3.1, Pydantic 2, PyTorch 2.6+, pandas 2.2+, Polars 1.x, ruff 0.9+, and uv 0.6+.

Recipe

import questionary
 
name = questionary.text("Project name:").ask()
framework = questionary.select("Framework?", choices=["FastAPI", "Django", "Flask"]).ask()
confirmed = questionary.confirm("Proceed?").ask()

When to reach for this:

Project scaffolding wizards (cookiecutter-style)
Destructive operations needing confirmation
Multi-step setup where flags would be unwieldy

Working Example

import questionary
from rich.console import Console
 
console = Console()
 
def run_setup_wizard() -> dict:
    if not questionary.confirm("Run interactive setup?").ask():
        raise SystemExit("Aborted")
 
    answers = {
        "name": questionary.text("Project name:", validate=lambda t: len(t) > 0 or "Required").ask(),
        "framework": questionary.select(
            "Web framework:",
            choices=["FastAPI", "Django", "Flask"],
        ).ask(),
        "database": questionary.checkbox(
            "Features:",
            choices=["postgres", "redis", "celery"],
        ).ask(),
        "deploy": questionary.select(
            "Deployment target:",
            choices=["Docker", "Lambda", "VPS"],
        ).ask(),
    }
 
    if not questionary.confirm(f"Create {answers['name']} with {answers['framework']}?").ask():
        raise SystemExit("Aborted")
 
    return answers
 
if __name__ == "__main__":
    config = run_setup_wizard()
    console.print(f"[green]Creating project: {config}[/green]")

What this demonstrates:

Text, select, checkbox, and confirm prompt types
Validation on text input
Abort on cancellation (Ctrl+C or No)
Rich output after wizard completes

Deep Dive

Prompt Selection

Library	Best for
`click.confirm/prompt`	Simple confirm/text
`questionary`	Select, checkbox, autocomplete
`prompt_toolkit`	Full custom REPL
`InquirerPy`	Alternative to questionary

Gotchas

Interactive prompts in CI - hangs forever. Fix: detect CI=true or --yes flag to skip prompts.
No default for destructive confirm - accidental yes. Fix: default=False on confirm().
Password echo - visible on screen. Fix: questionary.password() or click.prompt(hide_input=True).
Prompts without non-interactive fallback - breaks scripting. Fix: every prompt needs a CLI flag equivalent.
Unicode on Windows terminals - display issues. Fix: Rich Console or enable UTF-8 terminal.

Alternatives

Alternative	Use When	Don't Use When
CLI flags only	Scriptable automation	Guided first-time setup
web-based setup	Complex configuration	Terminal-only tool
config file	Repeatable deploys	Initial onboarding

FAQs

questionary or click.prompt?

click.prompt for simple text/confirm. questionary for select, checkbox, autocomplete.

How do I skip prompts in CI?

if os.environ.get("CI"): use defaults or --yes / --non-interactive flag.

How do I handle Ctrl+C?

questionary returns None on cancel. Check and exit gracefully.

Password input?

questionary.password("API key:").ask().

Autocomplete?

questionary.autocomplete("Path:", choices=file_list).ask().

Can I style prompts?

questionary supports style dictionaries for colors.

Testing interactive CLIs?

Mock questionary.text(...).ask() return values in pytest.

Multi-page wizards?

Chain prompts in functions; show progress with Rich.

prompt_toolkit when?

Building a full REPL or syntax-highlighted input. Heavier than questionary.

Default selections?

questionary.select(..., default="FastAPI").ask().

Click - built-in confirm/prompt
Rich Output - formatted wizard output
Config & Environment - post-wizard config
CLI Best Practices - interactive vs scriptable

Stack versions: This page was written for Python 3.14.0, FastAPI 0.115+, Django 5.2, Flask 3.1, Pydantic 2, PyTorch 2.6+, pandas 2.2+, Polars 1.x, ruff 0.9+, and uv 0.6+.

Interactive Prompts

Search Documentation

Interactive Prompts