Open-source AI coworkers that each get a computer of their own: a browser, files and tools, with every action decided before it happens and recorded after. Bring any AG-UI agent.
import { expect, test } from "bun:test";
import {
isNaming,
TYPING_MAX_SECONDS,
typingSeconds,
} from "../src/lib/typed-reveal";
test("a line that gains a name is a naming", () => {
expect(isNaming(undefined, "Travel receipt rules")).toBe(true);
});
test("a row that arrives already named is not", () => {
// Every row on every load takes this path. Treating it as an arrival is what would set the whole
// roster typing at once when somebody opens the app.
expect(isNaming("Travel receipt rules", "Travel receipt rules")).toBe(false);
});
test("a name replaced by another name is not an arrival either", () => {
// Nothing does this today. If something ever does, replaying the arrival would misdescribe it.
expect(isNaming("An older subject", "A newer subject")).toBe(false);
});
test("losing a name is not an arrival", () => {
expect(isNaming("Travel receipt rules", undefined)).toBe(false);
});
test("longer names type for longer, at the same speed", () => {
// Constant motion: twice the characters, twice the time, until the ceiling.
expect(typingSeconds(20)).toBeCloseTo(typingSeconds(10) * 2, 5);
});
test("a very long name is capped rather than outstaying the moment", () => {
expect(typingSeconds(500)).toBe(TYPING_MAX_SECONDS);
});