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 { describe, expect, test } from "bun:test";
import { createChannelTitler } from "../src/channels/titler";
function respondWith(body: unknown, status = 200) {
const calls: { url: string; body: Record<string, unknown> }[] = [];
const fetchImpl = (async (url: string | URL, init?: RequestInit) => {
calls.push({
url: String(url),
body: JSON.parse(String(init?.body ?? "{}")),
});
return new Response(
typeof body === "string" ? body : JSON.stringify(body),
{ status },
);
}) as unknown as typeof fetch;
return { calls, fetchImpl };
}
describe("asking the model for a title", () => {
test("sends the excerpt to the deployment's own model", async () => {
const { calls, fetchImpl } = respondWith({
choices: [{ message: { content: "Travel receipt rules" } }],
});
const answer = await createChannelTitler({
model: "gpt-4.1-mini",
resolveApiKey: async () => "key-123",
environment: {},
fetchImpl,
})("Asked: which receipts count as travel?");
expect(answer).toBe("Travel receipt rules");
expect(calls[0]?.url).toBe("https://api.openai.com/v1/chat/completions");
expect(calls[0]?.body.model).toBe("gpt-4.1-mini");
});
test("asks the endpoint OPENAI_BASE_URL names, not OpenAI", async () => {
const { calls, fetchImpl } = respondWith({
choices: [{ message: { content: "Travel receipt rules" } }],
});
await createChannelTitler({
model: "openai/gpt-5.6-terra",
resolveApiKey: async () => "gateway-key",
environment: { OPENAI_BASE_URL: "https://gateway.internal/v1" },
fetchImpl,
})("Asked: which receipts count as travel?");
expect(calls.map((call) => call.url)).toEqual([
"https://gateway.internal/v1/chat/completions",
]);
});
test("reads OPENAI_BASE_URL from the process when no environment is passed", async () => {
const { calls, fetchImpl } = respondWith({
choices: [{ message: { content: "Travel receipt rules" } }],
});
const previous = process.env.OPENAI_BASE_URL;
process.env.OPENAI_BASE_URL = "http://localhost:4010";
try {
await createChannelTitler({
model: "gpt-4.1-mini",
resolveApiKey: async () => "gateway-key",
fetchImpl,
})("Asked: anything?");
} finally {
if (previous === undefined) delete process.env.OPENAI_BASE_URL;
else process.env.OPENAI_BASE_URL = previous;
}
expect(calls[0]?.url).toBe("http://localhost:4010/v1/chat/completions");
});
test("asks nothing at all when the deployment has no key", async () => {
const { calls, fetchImpl } = respondWith({});
const answer = await createChannelTitler({
model: "gpt-4.1-mini",
resolveApiKey: async () => null,
fetchImpl,
})("Asked: anything?");
// Not a failure. A deployment with no model configured does not name conversations, and must
// not spend a request finding that out on every pass.
expect(answer).toBeNull();
expect(calls).toHaveLength(0);
});
test("throws when the provider refuses, so the work is tried again", async () => {
const { fetchImpl } = respondWith("rate limited", 429);
await expect(
createChannelTitler({
model: "gpt-4.1-mini",
resolveApiKey: async () => "key-123",
fetchImpl,
})("Asked: anything?"),
).rejects.toThrow("429");
});
test("carries only one line of the provider's complaint", async () => {
const { fetchImpl } = respondWith(
"<html>\n <body>Too many requests</body>\n</html>",
503,
);
// The message ends up on the work item as the reason it was released, so a provider that
// answers with a page of HTML must not put that page in the database.
const thrown = await createChannelTitler({
model: "gpt-4.1-mini",
resolveApiKey: async () => "key-123",
fetchImpl,
})("Asked: anything?").catch((error: Error) => error.message);
expect(thrown).not.toContain("\n");
});
test("an empty completion is no answer rather than an empty title", async () => {
const { fetchImpl } = respondWith({
choices: [{ message: { content: " " } }],
});
const answer = await createChannelTitler({
model: "gpt-4.1-mini",
resolveApiKey: async () => "key-123",
fetchImpl,
})("Asked: anything?");
expect(answer).toBeNull();
});
});