Supercharge your AI agents with data from the web and beyond. Building the library for superintelligence. 🔥
TypeScript aiai-agentsai-crawlerai-scrapingai-searchcrawlerdata-extractionhtml-to-markdownllmmarkdownscraperscrapingweb-crawlerweb-dataweb-data-extractionweb-scraperweb-scrapingweb-searchwebscraping
import { describeIf, TEST_PRODUCTION, TEST_SUITE_WEBSITE } from "../lib";
import {
endpointFeedback,
endpointFeedbackRaw,
endpointFeedbackWithFailure,
expectMapToSucceed,
idmux,
Identity,
map,
scrapeTimeout,
searchRawFull,
} from "./lib";
import { writeFeedbackJob } from "../../../lib/feedback-job-store";
let identity: Identity;
let secondaryIdentity: Identity;
beforeAll(async () => {
identity = await idmux({
name: "endpoint-feedback",
concurrency: 100,
credits: 1000000,
});
secondaryIdentity = await idmux({
name: "endpoint-feedback-other",
concurrency: 100,
credits: 1000000,
});
}, 20000);
// Skipped in self-hosted mode: depends on the production job log tables,
// Autumn refunds, and the per-team daily refund cap.
describeIf(TEST_PRODUCTION)("Generic endpoint feedback tests", () => {
it("records map feedback, exposes the map job id, and makes duplicate submissions idempotent", async () => {
const raw = await map(
{ url: TEST_SUITE_WEBSITE, limit: 1, timeout: scrapeTimeout },
identity,
);
expectMapToSucceed(raw);
expect(typeof raw.body.id).toBe("string");
const first = await endpointFeedback(
{
endpoint: "map",
jobId: raw.body.id,
rating: "bad",
issues: ["missing_expected_url"],
note: "The map result did not include the canonical page I expected.",
},
identity,
);
expect(first.success).toBe(true);
expect(first.creditsRefunded).toBe(1);
expect(first.alreadySubmitted).toBeFalsy();
const second = await endpointFeedback(
{
endpoint: "map",
jobId: raw.body.id,
rating: "partial",
issues: ["still_missing_expected_url"],
note: "Submitting twice should not refund twice.",
},
identity,
);
expect(second.success).toBe(true);
expect(second.creditsRefunded).toBe(0);
expect(second.alreadySubmitted).toBe(true);
}, 120000);
it("rejects endpoint feedback for a job owned by another team", async () => {
const raw = await map(
{ url: TEST_SUITE_WEBSITE, limit: 1, timeout: scrapeTimeout },
identity,
);
expectMapToSucceed(raw);
const failed = await endpointFeedbackWithFailure(
{
endpoint: "map",
jobId: raw.body.id,
rating: "bad",
note: "This team should not be able to see the job.",
},
secondaryIdentity,
);
expect(failed.error.toLowerCase()).toContain("not found");
expect((failed as any).feedbackErrorCode).toBe("JOB_NOT_FOUND");
}, 120000);
it("applies search feedback validation on the generic endpoint", async () => {
const raw = await searchRawFull(
{ query: "firecrawl generic feedback", limit: 3 },
identity,
);
expect(raw.statusCode).toBe(200);
expect(typeof raw.body.id).toBe("string");
const failed = await endpointFeedbackRaw(
{
endpoint: "search",
jobId: raw.body.id,
rating: "good",
note: "A good search rating must name a valuable source.",
},
identity,
);
expect(failed.statusCode).toBe(400);
expect(failed.body.success).toBe(false);
expect(String(failed.body.error).toLowerCase()).toContain("invalid");
}, 90000);
it("rejects generic search feedback for failed search jobs", async () => {
const searchId =
"00000000-0000-7000-8000-" +
Math.floor(Math.random() * 1e12)
.toString(16)
.padStart(12, "0");
// The feedback policy reads the job's outcome from the feedback job
// store, so a failed search is a failed feedback job row.
expect(
await writeFeedbackJob({
endpoint: "search",
jobId: searchId,
requestId: searchId,
teamId: identity.teamId,
succeeded: false,
creditsBilled: 2,
zeroDataRetention: false,
}),
).toBe(true);
const failed = await endpointFeedbackRaw(
{
endpoint: "search",
jobId: searchId,
rating: "bad",
missingContent: [{ topic: "Results" }],
},
identity,
);
expect(failed.statusCode).toBe(409);
expect(failed.body.success).toBe(false);
expect(failed.body.feedbackErrorCode).toBe("SEARCH_FAILED");
}, 30000);
});