firecrawl-mcp 3.21.1 → 3.21.3

package/dist/index.js

@@ -1347,7 +1347,7 @@ async function monitorRequest(session, path2, init = {}) {
     const s = qs.toString();
     if (s) url += `?${s}`;
   }
-  const headers = { "X-Origin": "mcp" };
+  const headers = { "X-Origin": "mcp-fastmcp" };
   if (apiKey) headers.Authorization = `Bearer ${apiKey}`;
   if (init.body !== void 0) headers["Content-Type"] = "application/json";
   const response = await fetch(url, {
@@ -2333,6 +2333,7 @@ function createClient(apiKey) {
   return new FirecrawlApp(config);
 }
 var ORIGIN = "mcp-fastmcp";
+var ORIGIN_HEADERS2 = { "X-Origin": ORIGIN };
 var SAFE_MODE = process.env.CLOUD_SERVICE === "true";
 function getClient(session) {
   if (process.env.CLOUD_SERVICE === "true") {
@@ -2837,6 +2838,7 @@ async function keylessEligible(clientIp) {
       `${resolveApiBaseUrl()}/v2/keyless/eligibility`,
       {
         headers: {
+          ...ORIGIN_HEADERS2,
           "x-firecrawl-keyless-ip": clientIp,
           "x-firecrawl-keyless-secret": secret
         }
@@ -2858,6 +2860,7 @@ function isKeylessMode(session) {
 }
 async function keylessPost(path2, body, session) {
   const headers = {
+    ...ORIGIN_HEADERS2,
     "Content-Type": "application/json"
   };
   if (session?.keylessClientIp && process.env.KEYLESS_PROXY_SECRET) {
@@ -2877,6 +2880,61 @@ async function keylessPost(path2, body, session) {
   }
   return json;
 }
+async function getCrawlStatusWithOrigin(client, jobId) {
+  const res = await client.http.get(
+    `/v2/crawl/${encodeURIComponent(jobId)}`,
+    ORIGIN_HEADERS2
+  );
+  const body = res?.data ?? {};
+  const initialDocs = Array.isArray(body.data) ? body.data : [];
+  if (!body.next) {
+    return {
+      id: jobId,
+      status: body.status,
+      completed: body.completed ?? 0,
+      total: body.total ?? 0,
+      creditsUsed: body.creditsUsed,
+      expiresAt: body.expiresAt,
+      next: body.next ?? null,
+      data: initialDocs
+    };
+  }
+  const docs = initialDocs.slice();
+  let current = body.next;
+  while (current) {
+    const pageRes = await client.http.get(current, ORIGIN_HEADERS2);
+    const payload = pageRes?.data ?? {};
+    if (!payload.success) break;
+    const pageData = Array.isArray(payload.data) ? payload.data : payload.data?.pages || [];
+    docs.push(...pageData);
+    current = payload.next ?? (Array.isArray(payload.data) ? null : payload.data?.next) ?? null;
+  }
+  return {
+    id: jobId,
+    status: body.status,
+    completed: body.completed ?? 0,
+    total: body.total ?? 0,
+    creditsUsed: body.creditsUsed,
+    expiresAt: body.expiresAt,
+    next: null,
+    data: docs
+  };
+}
+async function waitForCrawlCompletionWithOrigin(client, jobId, pollInterval = 2, timeout) {
+  const startedAt = Date.now();
+  while (true) {
+    const status = await getCrawlStatusWithOrigin(client, jobId);
+    if (["completed", "failed", "cancelled"].includes(String(status.status ?? ""))) {
+      return status;
+    }
+    if (timeout != null && Date.now() - startedAt > timeout * 1e3) {
+      throw new Error(`Crawl job ${jobId} did not complete within ${timeout}s`);
+    }
+    await new Promise(
+      (resolve) => setTimeout(resolve, Math.max(1e3, pollInterval * 1e3))
+    );
+  }
+}
 var feedbackIssueSchema = z4.string().trim().min(1).max(80).regex(
   /^[a-z0-9][a-z0-9_-]*$/,
   "Issue codes must use lowercase letters, numbers, underscores, or hyphens"
@@ -3024,6 +3082,7 @@ Pass the \`searchId\` returned by \`firecrawl_search\` (the \`id\` field on the
       }
       if (querySuggestions) body.querySuggestions = querySuggestions;
       const headers = {
+        ...ORIGIN_HEADERS2,
         "Content-Type": "application/json"
       };
       const apiKey = session?.firecrawlApiKey;
@@ -3127,6 +3186,7 @@ Do not store multi-MB outputs in feedback. Use concise notes, issue codes, URLs,
       } = args2;
       const apiBase = resolveApiBaseUrl();
       const headers = {
+        ...ORIGIN_HEADERS2,
         "Content-Type": "application/json"
       };
       const apiKey = session?.firecrawlApiKey;
@@ -3250,11 +3310,26 @@ server.addTool({
     if (webhook) opts.webhook = webhook;
     delete opts.webhookHeaders;
     const cleaned = removeEmptyTopLevel(opts);
+    const pollInterval = typeof cleaned.pollInterval === "number" ? cleaned.pollInterval : 2;
+    const timeout = typeof cleaned.timeout === "number" ? cleaned.timeout : void 0;
+    delete cleaned.pollInterval;
+    delete cleaned.timeout;
     log.info("Starting crawl", { url: String(url) });
-    const res = await client.crawl(String(url), {
+    const started = await client.http.post("/v2/crawl", {
+      url: String(url),
       ...cleaned,
       origin: ORIGIN
     });
+    const crawlId = started?.data?.id;
+    if (!crawlId) {
+      return asText2(started?.data ?? {});
+    }
+    const res = await waitForCrawlCompletionWithOrigin(
+      client,
+      crawlId,
+      pollInterval,
+      timeout
+    );
     return asText2(res);
   }
 });
@@ -3286,7 +3361,8 @@ Check the status of a crawl job.
   parameters: z4.object({ id: z4.string() }),
   execute: async (args2, { session }) => {
     const client = getClient(session);
-    const res = await client.getCrawlStatus(args2.id);
+    const id = args2.id;
+    const res = await getCrawlStatusWithOrigin(client, id);
     return asText2(res);
   }
 });
@@ -3510,8 +3586,11 @@ Check the status of an agent job and retrieve results when complete. Use this to
     const client = getClient(session);
     const { id } = args2;
     log.info("Checking agent status", { id });
-    const res = await client.getAgentStatus(id);
-    return asText2(res);
+    const res = await client.http.get(
+      `/v2/agent/${encodeURIComponent(id)}`,
+      ORIGIN_HEADERS2
+    );
+    return asText2(res?.data ?? {});
   }
 });
 server.addTool({
@@ -3616,8 +3695,11 @@ Stop an interact session for a scraped page. Call this when you are done interac
     const client = getClient(session);
     const { scrapeId } = args2;
     log.info("Stopping interact session", { scrapeId });
-    const res = await client.stopInteraction(scrapeId);
-    return asText2(res);
+    const res = await client.http.delete(
+      `/v2/scrape/${encodeURIComponent(scrapeId)}/interact`,
+      ORIGIN_HEADERS2
+    );
+    return asText2(res?.data ?? {});
   }
 });
 if (process.env.CLOUD_SERVICE !== "true") {
@@ -3789,7 +3871,7 @@ Add \`"parsers": ["pdf"]\` (optionally with \`pdfOptions.maxPages\`) when parsin
       });
       form.append("file", blob, filename);
       form.append("options", JSON.stringify(optionsPayload));
-      const headers = {};
+      const headers = { ...ORIGIN_HEADERS2 };
       const apiKey = session?.firecrawlApiKey;
       if (apiKey) {
         headers["Authorization"] = `Bearer ${apiKey}`;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "firecrawl-mcp",
-  "version": "3.21.1",
+  "version": "3.21.3",
   "description": "MCP server for Firecrawl — search, scrape, and interact with the web. Supports both cloud and self-hosted instances. Features include web search, scraping, page interaction, batch processing, and LLM-powered content analysis.",
   "type": "module",
   "mcpName": "io.github.firecrawl/firecrawl-mcp-server",