An .angl chapter is English behavior plus executable examples. A compiler agent turns it into ordinary code. A black-box judge accepts the generated edition only if every example passes.
# Fetch Price
> Boundary: `fetch_price(url: string) -> number`
> Runs as: `python`
## Purpose
Provide the price value from a JSON HTTP response.
## Behavior
Fetch JSON from the given URL and return the `price` field as a
number. If the field is missing, fail with an error that mentions
`price`.
## Examples
### The service returns a price
When the URL serves a price, the chapter returns that price.
Fixture `http_fixture`:
```json
{ "price": 19.99 }
```
Returns:
```json
19.99
```
### The service omits the price
When the URL serves JSON without a price field, the chapter
rejects it.
Fixture `http_fixture`:
```json
{}
```
Fails with error containing: priceOne composed program can cross Python, Node, Ruby, Go, Rust, and TypeScript, because each chapter picks its own target and every seam is contract-verified.
A real pydantic v1 to v2 break rotted a generated artifact. The unchanged chapter was recompiled under the new pin and passed the same contract. The durable asset is the spec, not the code.
Generated editions never grow hand-written test suites. If a behavior matters after regeneration, it becomes an example in the chapter, and every future edition has to pass it.
The docs cover the chapter format, the judge, fixtures, composition, compile targets, and the honest limits.