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

# Quickstart

> Install the Sense interpreter and run your first program in about two minutes.

## Install

```bash theme={null}
pip install sense-lang
```

Requires Python 3.10+. This installs the `sense` command along with the
interpreter itself — no other runtime dependency.

## Run a file

Create `hello.sns`:

```sns hello.sns theme={null}
print("hello, sense")
```

Run it:

```bash theme={null}
sense run hello.sns
```

```txt theme={null}
hello, sense
```

## Try the REPL

```bash theme={null}
sense repl
# or just: sense
```

```txt theme={null}
Sense 0.20.0 — interactive REPL ('exit' to quit)
sense> x = 5
sense> print(x * 2)
10
sense> exit
```

Typing `sense` with no arguments does the same thing — the same
convenience bare `python` has (`sense --help` still lists every
subcommand).

<Info>
  Sense blocks are indentation, like Python's — the REPL keeps prompting with
  `...` while a block (anything ending in `:`) is open, and a blank line ends
  it. See [CLI Reference](/reference/cli) for the full behavior.
</Info>

## A slightly bigger example

```sns fib.sns theme={null}
def fib(n) -> Int:
    if n < 2:
        return n
    return fib(n - 1) + fib(n - 2)

i = 0
while i < 10:
    print(fib(i))
    i = i + 1
```

```bash theme={null}
sense run fib.sns
```

```txt theme={null}
0
1
1
2
3
5
8
13
21
34
```

## The AI-native part

Everything above is Sense's deterministic core — no different in spirit
from Python. Here's the part that isn't: `ask()`, running entirely
offline against a mock model so this needs no API key.

```sns reason_demo.sns theme={null}
model reasoning = inference("mock", "demo", response_template: "I think: {prompt}")
set delegation = reasoning

result = ask("what database should we use?")

print(result.value)        # the model's answer
print(result.confidence)   # how sure it was, as a plain readable number
```

```bash theme={null}
sense run reason_demo.sns
```

```txt theme={null}
I think: what database should we use?
0.5
```

`result` is an `Answer`, not a plain string — accessing `.value` is
mandatory, so there's never a moment where AI-generated output silently
looks identical to a fact your code computed itself. See
[Model, Ask & Answer](/ai-native/model-ask-answer) for the full
picture, including how to scope which model is active with `session`.

## Next

<CardGroup cols={2}>
  <Card title="Tutorials" icon="graduation-cap" href="/tutorials/ask-a-model">
    Three short, hands-on walkthroughs — start with "Ask a Model."
  </Card>

  <Card title="Which keyword do I want?" icon="list-check" href="/reference/keywords#which-keyword-do-i-want">
    `def` vs `tool` vs `action` vs `agent` — a one-glance decision guide.
  </Card>

  <Card title="Language guide" icon="book-open" href="/language/values-and-types">
    Values, scoping, functions, control flow, modules.
  </Card>

  <Card title="Actions & policy" icon="shield-check" href="/safety/actions">
    How Sense makes world-changing operations safe by construction.
  </Card>
</CardGroup>