firecrawl 4.42.0 → 4.42.2

package/README.md

@@ -102,6 +102,13 @@ const parsed = await app.parse(
 console.log(parsed.markdown);
 ```
 
+To see which file formats `parse` accepts, call `getParseFormats`. Each entry has `format`, `kind` (for example, `document` or `image`), `extensions`, `mimeTypes`, and `available` (false when the format is not enabled on this deployment).
+
+```js
+const formats = await app.getParseFormats();
+console.log(formats.filter((f) => f.available).map((f) => f.format));
+```
+
 ### Crawling a Website
 
 To crawl a website with error handling, use the `crawl` method. It takes the starting URL and optional parameters, including limits and per‑page `scrapeOptions`.

package/dist/{chunk-OCJTYRX5.js → chunk-XZ5XI5DK.js}

@@ -12,7 +12,7 @@ var require_package = __commonJS({
   "package.json"(exports, module) {
     module.exports = {
       name: "@mendable/firecrawl-js",
-      version: "4.42.0",
+      version: "4.42.2",
       description: "JavaScript SDK for the Firecrawl API: web scraping, crawling, web search, and scientific literature search over a research paper index of PubMed, bioRxiv, medRxiv and arXiv abstracts",
       main: "dist/index.js",
       types: "dist/index.d.ts",
@@ -38,7 +38,7 @@ var require_package = __commonJS({
       author: "Mendable.ai",
       license: "MIT",
       dependencies: {
-        axios: "1.18.0",
+        axios: "1.20.0",
         firecrawl: "4.30.1",
         "typescript-event-target": "^1.1.1",
         zod: "^3.23.8",

package/dist/index.cjs

@@ -39,7 +39,7 @@ var require_package = __commonJS({
   "package.json"(exports2, module2) {
     module2.exports = {
       name: "@mendable/firecrawl-js",
-      version: "4.42.0",
+      version: "4.42.2",
       description: "JavaScript SDK for the Firecrawl API: web scraping, crawling, web search, and scientific literature search over a research paper index of PubMed, bioRxiv, medRxiv and arXiv abstracts",
       main: "dist/index.js",
       types: "dist/index.d.ts",
@@ -65,7 +65,7 @@ var require_package = __commonJS({
       author: "Mendable.ai",
       license: "MIT",
       dependencies: {
-        axios: "1.18.0",
+        axios: "1.20.0",
         firecrawl: "4.30.1",
         "typescript-event-target": "^1.1.1",
         zod: "^3.23.8",
@@ -150,6 +150,21 @@ function getVersion() {
   }
 }
 
+// src/utils/apiOrigin.ts
+function pinToApiOrigin(apiUrl, url) {
+  const normalized = url.replace(/[\t\n\r]/g, "").replace(/^[\u0000-\u0020]+/, "");
+  if (!/^([a-z][a-z\d+\-.]*:|[\\/]{2})/i.test(normalized)) return url;
+  const base = URL.canParse(apiUrl) ? new URL(apiUrl) : null;
+  if (!base?.host) {
+    throw new Error(`apiUrl must be an absolute URL, got ${JSON.stringify(apiUrl)}`);
+  }
+  const target = new URL(normalized, base);
+  const pinned = new URL(base.origin);
+  pinned.pathname = target.pathname;
+  pinned.search = target.search;
+  return pinned.href;
+}
+
 // src/v2/utils/httpClient.ts
 var HttpClient = class {
   instance;
@@ -190,6 +205,7 @@ var HttpClient = class {
     config.headers = {
       ...config.headers || {}
     };
+    if (config.url) config.url = pinToApiOrigin(this.apiUrl, config.url);
     let lastError;
     for (let attempt = 0; attempt < this.maxRetries; attempt++) {
       try {
@@ -770,6 +786,16 @@ async function parse(http, file, options) {
     throw err;
   }
 }
+async function getParseFormats(http) {
+  try {
+    const res = await http.get("/v2/parse/formats");
+    if (res.status !== 200 || !res.data?.success) throwForBadResponse(res, "get parse formats");
+    return res.data.data?.formats ?? [];
+  } catch (err) {
+    if (err?.isAxiosError) return normalizeAxiosError(err, "get parse formats");
+    throw err;
+  }
+}
 
 // src/v2/methods/search.ts
 function prepareSearchPayload(req) {
@@ -2352,6 +2378,13 @@ var FirecrawlClient = class {
   async parse(file, options) {
     return parse(this.http, file, options);
   }
+  /**
+   * List the file formats the parse endpoint accepts on this deployment.
+   * @returns Formats with their kind, extensions, MIME types, and availability.
+   */
+  async getParseFormats() {
+    return getParseFormats(this.http);
+  }
   // Search
   /**
    * Search the web and optionally scrape each result.
@@ -3020,7 +3053,7 @@ var FirecrawlApp = class {
       throw new FirecrawlError("No crawl ID provided", 400);
     }
     const headers = this.prepareHeaders();
-    const targetURL = new URL(nextURL ?? `${this.apiUrl}/v1/crawl/${id}`);
+    const targetURL = new URL(nextURL ?? `${this.apiUrl}/v1/crawl/${id}`, this.apiUrl);
     if (skip !== void 0) {
       targetURL.searchParams.set("skip", skip.toString());
     }
@@ -3265,7 +3298,7 @@ var FirecrawlApp = class {
       throw new FirecrawlError("No batch scrape ID provided", 400);
     }
     const headers = this.prepareHeaders();
-    const targetURL = new URL(nextURL ?? `${this.apiUrl}/v1/batch/scrape/${id}`);
+    const targetURL = new URL(nextURL ?? `${this.apiUrl}/v1/batch/scrape/${id}`, this.apiUrl);
     if (skip !== void 0) {
       targetURL.searchParams.set("skip", skip.toString());
     }
@@ -3477,7 +3510,7 @@ var FirecrawlApp = class {
    */
   async getRequest(url, headers) {
     try {
-      return await import_axios3.default.get(url, { headers });
+      return await import_axios3.default.get(pinToApiOrigin(this.apiUrl, url), { headers });
     } catch (error) {
       if (error instanceof import_axios3.AxiosError && error.response) {
         return error.response;

package/dist/index.d.cts

@@ -267,6 +267,17 @@ type ParseOptions = Omit<ScrapeOptions, "formats" | "waitFor" | "mobile" | "acti
     formats?: ParseFormatOption[];
     proxy?: "basic" | "auto";
 };
+/** A file format accepted by the parse endpoint. */
+interface ParseFormatInfo {
+    /** Format identifier, e.g. "pdf". */
+    format: string;
+    kind: "document" | "image" | (string & {});
+    /** File extensions including the leading dot, e.g. ".pdf". */
+    extensions: string[];
+    mimeTypes: string[];
+    /** False when the format is recognized but not enabled on this deployment. */
+    available: boolean;
+}
 interface WebhookConfig {
     url: string;
     headers?: Record<string, string>;
@@ -1409,6 +1420,11 @@ interface AgentStatusResponse {
     status: "processing" | "completed" | "failed";
     error?: string;
     data?: unknown;
+    /** Best-effort result on a failed run; never a completed `data` value. */
+    partial?: unknown;
+    /** Schema validity of `partial`, when a schema was supplied. */
+    partialSchemaValid?: boolean;
+    stopReason?: "credit_limit_reached";
     /**
      * Server-provided model name. Widened past the request-side union on
      * purpose: new models ship without an SDK release, so pinning this to known
@@ -1447,6 +1463,9 @@ interface AgentThreadRun {
     message: string | null;
     /** Only present when the request asked for includeData. */
     data?: unknown;
+    partial?: unknown;
+    partialSchemaValid?: boolean;
+    stopReason?: "credit_limit_reached";
     suggestions?: AgentSuggestion[] | null;
     pendingApproval?: PendingApproval | null;
     exchange?: AgentExchangeSummary | null;
@@ -1676,6 +1695,8 @@ interface CrawlErrorsResponse$1 {
         url: string;
         code?: string;
         error: string;
+        /** Present when the page needs provider terms accepted first. */
+        requiresAction?: RequiresAction;
     }[];
     robotsBlocked: string[];
 }
@@ -1997,11 +2018,11 @@ declare class HttpClient {
     getTimeoutMs(): number;
     private request;
     private sleep;
-    post<T = any>(endpoint: string, body: Record<string, unknown>, options?: RequestOptions): Promise<AxiosResponse<T, any, {}>>;
-    postMultipart<T = any>(endpoint: string, formData: FormData, options?: RequestOptions): Promise<AxiosResponse<T, any, {}>>;
-    get<T = any>(endpoint: string, headers?: Record<string, string>): Promise<AxiosResponse<T, any, {}>>;
-    delete<T = any>(endpoint: string, headers?: Record<string, string>): Promise<AxiosResponse<T, any, {}>>;
-    patch<T = any>(endpoint: string, body: Record<string, unknown>, options?: RequestOptions): Promise<AxiosResponse<T, any, {}>>;
+    post<T = any>(endpoint: string, body: Record<string, unknown>, options?: RequestOptions): Promise<AxiosResponse<T, any, {}, any>>;
+    postMultipart<T = any>(endpoint: string, formData: FormData, options?: RequestOptions): Promise<AxiosResponse<T, any, {}, any>>;
+    get<T = any>(endpoint: string, headers?: Record<string, string>): Promise<AxiosResponse<T, any, {}, any>>;
+    delete<T = any>(endpoint: string, headers?: Record<string, string>): Promise<AxiosResponse<T, any, {}, any>>;
+    patch<T = any>(endpoint: string, body: Record<string, unknown>, options?: RequestOptions): Promise<AxiosResponse<T, any, {}, any>>;
     prepareHeaders(idempotencyKey?: string): Record<string, string>;
 }
 
@@ -2056,6 +2077,7 @@ declare function prepareAgentPayload(args: {
     origin?: string;
     maxCredits?: number;
     strictConstrainToURLs?: boolean;
+    /** Defaults to spark-2. "spark-1-pro" and "spark-1-mini" are deprecated and run spark-2. */
     model?: "spark-1-pro" | "spark-1-mini" | "spark-2";
     effort?: AgentEffort;
     webhook?: string | AgentWebhookConfig;
@@ -2282,6 +2304,11 @@ declare class FirecrawlClient {
         json?: InferredJsonFromOptions<Opts>;
     }>;
     parse(file: ParseFile, options?: ParseOptions): Promise<Document>;
+    /**
+     * List the file formats the parse endpoint accepts on this deployment.
+     * @returns Formats with their kind, extensions, MIME types, and availability.
+     */
+    getParseFormats(): Promise<ParseFormatInfo[]>;
     /**
      * Search the web and optionally scrape each result.
      * @param query Search query string.
@@ -2993,6 +3020,8 @@ interface CrawlErrorsResponse {
         url: string;
         code?: string;
         error: string;
+        /** Present when the page needs provider terms accepted first. */
+        requiresAction?: RequiresAction;
     }[];
     /**
      * URLs blocked by robots.txt
@@ -3528,4 +3557,4 @@ declare class Firecrawl extends FirecrawlClient {
     get v1(): FirecrawlApp;
 }
 
-export { type ActionOption, type ActiveCrawl, type ActiveCrawlsResponse, type AgentEffort, type AgentExchangeOptions, type AgentExchangeSummary, type AgentListItem, type AgentListOptions, type AgentListResponse, type AgentMode, type AgentOnTermsRequired, type AgentOptions$1 as AgentOptions, type AgentResponse, type AgentSkippedProvider, type AgentSnapshotResponse, type AgentStatusResponse, type AgentSuggestion, type AgentTermsGate, type AgentTermsRequiredAction, type AgentThread, type AgentThreadResponse, type AgentThreadRun, type AgentTraceActiveBrowserSession, type AgentTraceAgentFinishedEvent, type AgentTraceAgentIdentity, type AgentTraceAgentRole, type AgentTraceAgentStartedEvent, type AgentTraceArtifactChange, type AgentTraceArtifactUpdatedEvent, type AgentTraceBrowserSessionFinishedEvent, type AgentTraceBrowserSessionStartedEvent, type AgentTraceError, type AgentTraceErrorOccurredEvent, type AgentTraceEvent, type AgentTraceProgressReportedEvent, type AgentTraceReasoningSummaryEvent, type AgentTraceResponse, type AgentTraceRunCancelRequestedEvent, type AgentTraceRunFinishedEvent, type AgentTraceRunStartedEvent, type AgentTraceToolCallFinishedEvent, type AgentTraceToolCallStartedEvent, type AgentWebhookConfig, type AgentWebhookEvent, type AlexandriaCall, type AlexandriaOptions, type AlexandriaScrapeData, type AlexandriaScrapeError, type AlexandriaScrapeRequest, type AlexandriaScrapeResult, type AttributesFormat, type AuditMetadata$1 as AuditMetadata, type BatchScrapeJob, type BatchScrapeOptions, type BatchScrapeResponse$1 as BatchScrapeResponse, type BrandingProfile, type BrowserCreateResponse, type BrowserDeleteResponse, type BrowserExecuteResponse, type BrowserListResponse, type BrowserSession, type CategoryOption, type ChangeTrackingFormat, type ClickAction, type ConcurrencyCheck, type CrawlErrorsResponse$1 as CrawlErrorsResponse, type CrawlJob, type CrawlOptions, type CrawlResponse$1 as CrawlResponse, type CreateMonitorRequest, type CreditUsage, type CreditUsageHistoricalPeriod, type CreditUsageHistoricalResponse, type DeveloperSearchLicenseDisclosure, type DeveloperSearchOptions, type DeveloperSearchPassage, type DeveloperSearchRepoStatus, type DeveloperSearchResponse, type DeveloperSearchResult, type DeveloperSearchSourceStatus, type DeveloperSearchType, type DiscoveredTool, type Document, type DocumentMetadata, type EndpointFeedbackEndpoint, type EndpointFeedbackRequest, type ErrorDetails, type ExecuteJavascriptAction, type ExtractResponse$1 as ExtractResponse, type FeedbackMissingContent, type FeedbackRating, type FeedbackResponse, type FeedbackValuableSource, type FindToolsData, type FindToolsOptions, Firecrawl, FirecrawlApp as FirecrawlAppV1, FirecrawlClient, type FirecrawlClientInput, type FirecrawlClientOptions, type Format, type FormatOption, type FormatString, type GetMonitorCheckOptions, type GetPaperOptions, type GitHubScoreBreakdown, type GitHubSearchItem, type GitHubSearchResponse, type HighlightsFormat, type IdMap, type ImageParser, JobTimeoutError, type JsonFormat, type ListMonitorChecksOptions, type ListMonitorsOptions, type LocationConfig$1 as LocationConfig, type MapData, type MapOptions, type MenuAvailability, type MenuImage, type MenuItem, type MenuItemIdentifiers, type MenuMerchant, type MenuPrice, type MenuProfile, type MenuSection, type Monitor, type MonitorCheck, type MonitorCheckDetail, type MonitorCheckPage, type MonitorCrawlTarget, type MonitorCrawlTargetResult, type MonitorEmailNotification, type MonitorEmailRecipientSubscription, type MonitorJsonFieldDiff, type MonitorNotification, type MonitorPageDiff, type MonitorPageJudgment, type MonitorPageSnapshot, type MonitorSchedule, type MonitorScrapeTarget, type MonitorScrapeTargetResult, type MonitorSearchTarget, type MonitorSearchTargetResult, type MonitorSummary, type MonitorTarget, type MonitorTargetResult, type MonitorWebhookConfig, type PDFAction, type PDFParser, type PaginationConfig, type PaperMetadata, type PaperMetadataResponse, type PaperResult, type PaperSignals, type ParseFile, type ParseFileData, type ParseFormat, type ParseFormatOption, type ParseFormatString, type ParseOptions, type Passage, type PdfBlockConfidence, type PdfBlockItem, type PdfPage, type PdfPageBlocks, type PendingApproval, type PendingApprovalCall, type PendingCallsApproval, type PendingTermsApproval, type PressAction, type ProductAvailability, type ProductImage, type ProductPrice, type ProductProfile, type ProductSale, type ProductVariant, type QueryFormat, type QuestionFormat, type QueueStatusResponse$1 as QueueStatusResponse, type ReadPaperResponse, type RedactPIIEntity, type RedactPIIOptions, type RequiresAction, ResearchClient, type ScrapeAction, type ScrapeBrowserDeleteResponse, type ScrapeExecuteRequest, type ScrapeExecuteResponse, type ScrapeOptions, type ScreenshotAction, type ScreenshotFormat, type ScrollAction, SdkError, type SearchData, type SearchFeedbackRequest, type SearchGithubOptions, type SearchPapersOptions, type SearchPapersResponse, type SearchRequest, type SearchResultImages, type SearchResultNews, type SearchResultWeb, type SimilarPapersOptions, type SimilarPapersResponse, type ThreatProtectionOptions, type TokenUsage, type TokenUsageHistoricalPeriod, type TokenUsageHistoricalResponse, type UpdateMonitorRequest, type Viewport, type WaitAction, Watcher, type WatcherOptions, type WebhookConfig, type WriteAction, Firecrawl as default, parseRequiresAction };
+export { type ActionOption, type ActiveCrawl, type ActiveCrawlsResponse, type AgentEffort, type AgentExchangeOptions, type AgentExchangeSummary, type AgentListItem, type AgentListOptions, type AgentListResponse, type AgentMode, type AgentOnTermsRequired, type AgentOptions$1 as AgentOptions, type AgentResponse, type AgentSkippedProvider, type AgentSnapshotResponse, type AgentStatusResponse, type AgentSuggestion, type AgentTermsGate, type AgentTermsRequiredAction, type AgentThread, type AgentThreadResponse, type AgentThreadRun, type AgentTraceActiveBrowserSession, type AgentTraceAgentFinishedEvent, type AgentTraceAgentIdentity, type AgentTraceAgentRole, type AgentTraceAgentStartedEvent, type AgentTraceArtifactChange, type AgentTraceArtifactUpdatedEvent, type AgentTraceBrowserSessionFinishedEvent, type AgentTraceBrowserSessionStartedEvent, type AgentTraceError, type AgentTraceErrorOccurredEvent, type AgentTraceEvent, type AgentTraceProgressReportedEvent, type AgentTraceReasoningSummaryEvent, type AgentTraceResponse, type AgentTraceRunCancelRequestedEvent, type AgentTraceRunFinishedEvent, type AgentTraceRunStartedEvent, type AgentTraceToolCallFinishedEvent, type AgentTraceToolCallStartedEvent, type AgentWebhookConfig, type AgentWebhookEvent, type AlexandriaCall, type AlexandriaOptions, type AlexandriaScrapeData, type AlexandriaScrapeError, type AlexandriaScrapeRequest, type AlexandriaScrapeResult, type AttributesFormat, type AuditMetadata$1 as AuditMetadata, type BatchScrapeJob, type BatchScrapeOptions, type BatchScrapeResponse$1 as BatchScrapeResponse, type BrandingProfile, type BrowserCreateResponse, type BrowserDeleteResponse, type BrowserExecuteResponse, type BrowserListResponse, type BrowserSession, type CategoryOption, type ChangeTrackingFormat, type ClickAction, type ConcurrencyCheck, type CrawlErrorsResponse$1 as CrawlErrorsResponse, type CrawlJob, type CrawlOptions, type CrawlResponse$1 as CrawlResponse, type CreateMonitorRequest, type CreditUsage, type CreditUsageHistoricalPeriod, type CreditUsageHistoricalResponse, type DeveloperSearchLicenseDisclosure, type DeveloperSearchOptions, type DeveloperSearchPassage, type DeveloperSearchRepoStatus, type DeveloperSearchResponse, type DeveloperSearchResult, type DeveloperSearchSourceStatus, type DeveloperSearchType, type DiscoveredTool, type Document, type DocumentMetadata, type EndpointFeedbackEndpoint, type EndpointFeedbackRequest, type ErrorDetails, type ExecuteJavascriptAction, type ExtractResponse$1 as ExtractResponse, type FeedbackMissingContent, type FeedbackRating, type FeedbackResponse, type FeedbackValuableSource, type FindToolsData, type FindToolsOptions, Firecrawl, FirecrawlApp as FirecrawlAppV1, FirecrawlClient, type FirecrawlClientInput, type FirecrawlClientOptions, type Format, type FormatOption, type FormatString, type GetMonitorCheckOptions, type GetPaperOptions, type GitHubScoreBreakdown, type GitHubSearchItem, type GitHubSearchResponse, type HighlightsFormat, type IdMap, type ImageParser, JobTimeoutError, type JsonFormat, type ListMonitorChecksOptions, type ListMonitorsOptions, type LocationConfig$1 as LocationConfig, type MapData, type MapOptions, type MenuAvailability, type MenuImage, type MenuItem, type MenuItemIdentifiers, type MenuMerchant, type MenuPrice, type MenuProfile, type MenuSection, type Monitor, type MonitorCheck, type MonitorCheckDetail, type MonitorCheckPage, type MonitorCrawlTarget, type MonitorCrawlTargetResult, type MonitorEmailNotification, type MonitorEmailRecipientSubscription, type MonitorJsonFieldDiff, type MonitorNotification, type MonitorPageDiff, type MonitorPageJudgment, type MonitorPageSnapshot, type MonitorSchedule, type MonitorScrapeTarget, type MonitorScrapeTargetResult, type MonitorSearchTarget, type MonitorSearchTargetResult, type MonitorSummary, type MonitorTarget, type MonitorTargetResult, type MonitorWebhookConfig, type PDFAction, type PDFParser, type PaginationConfig, type PaperMetadata, type PaperMetadataResponse, type PaperResult, type PaperSignals, type ParseFile, type ParseFileData, type ParseFormat, type ParseFormatInfo, type ParseFormatOption, type ParseFormatString, type ParseOptions, type Passage, type PdfBlockConfidence, type PdfBlockItem, type PdfPage, type PdfPageBlocks, type PendingApproval, type PendingApprovalCall, type PendingCallsApproval, type PendingTermsApproval, type PressAction, type ProductAvailability, type ProductImage, type ProductPrice, type ProductProfile, type ProductSale, type ProductVariant, type QueryFormat, type QuestionFormat, type QueueStatusResponse$1 as QueueStatusResponse, type ReadPaperResponse, type RedactPIIEntity, type RedactPIIOptions, type RequiresAction, ResearchClient, type ScrapeAction, type ScrapeBrowserDeleteResponse, type ScrapeExecuteRequest, type ScrapeExecuteResponse, type ScrapeOptions, type ScreenshotAction, type ScreenshotFormat, type ScrollAction, SdkError, type SearchData, type SearchFeedbackRequest, type SearchGithubOptions, type SearchPapersOptions, type SearchPapersResponse, type SearchRequest, type SearchResultImages, type SearchResultNews, type SearchResultWeb, type SimilarPapersOptions, type SimilarPapersResponse, type ThreatProtectionOptions, type TokenUsage, type TokenUsageHistoricalPeriod, type TokenUsageHistoricalResponse, type UpdateMonitorRequest, type Viewport, type WaitAction, Watcher, type WatcherOptions, type WebhookConfig, type WriteAction, Firecrawl as default, parseRequiresAction };

package/dist/index.d.ts

@@ -267,6 +267,17 @@ type ParseOptions = Omit<ScrapeOptions, "formats" | "waitFor" | "mobile" | "acti
     formats?: ParseFormatOption[];
     proxy?: "basic" | "auto";
 };
+/** A file format accepted by the parse endpoint. */
+interface ParseFormatInfo {
+    /** Format identifier, e.g. "pdf". */
+    format: string;
+    kind: "document" | "image" | (string & {});
+    /** File extensions including the leading dot, e.g. ".pdf". */
+    extensions: string[];
+    mimeTypes: string[];
+    /** False when the format is recognized but not enabled on this deployment. */
+    available: boolean;
+}
 interface WebhookConfig {
     url: string;
     headers?: Record<string, string>;
@@ -1409,6 +1420,11 @@ interface AgentStatusResponse {
     status: "processing" | "completed" | "failed";
     error?: string;
     data?: unknown;
+    /** Best-effort result on a failed run; never a completed `data` value. */
+    partial?: unknown;
+    /** Schema validity of `partial`, when a schema was supplied. */
+    partialSchemaValid?: boolean;
+    stopReason?: "credit_limit_reached";
     /**
      * Server-provided model name. Widened past the request-side union on
      * purpose: new models ship without an SDK release, so pinning this to known
@@ -1447,6 +1463,9 @@ interface AgentThreadRun {
     message: string | null;
     /** Only present when the request asked for includeData. */
     data?: unknown;
+    partial?: unknown;
+    partialSchemaValid?: boolean;
+    stopReason?: "credit_limit_reached";
     suggestions?: AgentSuggestion[] | null;
     pendingApproval?: PendingApproval | null;
     exchange?: AgentExchangeSummary | null;
@@ -1676,6 +1695,8 @@ interface CrawlErrorsResponse$1 {
         url: string;
         code?: string;
         error: string;
+        /** Present when the page needs provider terms accepted first. */
+        requiresAction?: RequiresAction;
     }[];
     robotsBlocked: string[];
 }
@@ -1997,11 +2018,11 @@ declare class HttpClient {
     getTimeoutMs(): number;
     private request;
     private sleep;
-    post<T = any>(endpoint: string, body: Record<string, unknown>, options?: RequestOptions): Promise<AxiosResponse<T, any, {}>>;
-    postMultipart<T = any>(endpoint: string, formData: FormData, options?: RequestOptions): Promise<AxiosResponse<T, any, {}>>;
-    get<T = any>(endpoint: string, headers?: Record<string, string>): Promise<AxiosResponse<T, any, {}>>;
-    delete<T = any>(endpoint: string, headers?: Record<string, string>): Promise<AxiosResponse<T, any, {}>>;
-    patch<T = any>(endpoint: string, body: Record<string, unknown>, options?: RequestOptions): Promise<AxiosResponse<T, any, {}>>;
+    post<T = any>(endpoint: string, body: Record<string, unknown>, options?: RequestOptions): Promise<AxiosResponse<T, any, {}, any>>;
+    postMultipart<T = any>(endpoint: string, formData: FormData, options?: RequestOptions): Promise<AxiosResponse<T, any, {}, any>>;
+    get<T = any>(endpoint: string, headers?: Record<string, string>): Promise<AxiosResponse<T, any, {}, any>>;
+    delete<T = any>(endpoint: string, headers?: Record<string, string>): Promise<AxiosResponse<T, any, {}, any>>;
+    patch<T = any>(endpoint: string, body: Record<string, unknown>, options?: RequestOptions): Promise<AxiosResponse<T, any, {}, any>>;
     prepareHeaders(idempotencyKey?: string): Record<string, string>;
 }
 
@@ -2056,6 +2077,7 @@ declare function prepareAgentPayload(args: {
     origin?: string;
     maxCredits?: number;
     strictConstrainToURLs?: boolean;
+    /** Defaults to spark-2. "spark-1-pro" and "spark-1-mini" are deprecated and run spark-2. */
     model?: "spark-1-pro" | "spark-1-mini" | "spark-2";
     effort?: AgentEffort;
     webhook?: string | AgentWebhookConfig;
@@ -2282,6 +2304,11 @@ declare class FirecrawlClient {
         json?: InferredJsonFromOptions<Opts>;
     }>;
     parse(file: ParseFile, options?: ParseOptions): Promise<Document>;
+    /**
+     * List the file formats the parse endpoint accepts on this deployment.
+     * @returns Formats with their kind, extensions, MIME types, and availability.
+     */
+    getParseFormats(): Promise<ParseFormatInfo[]>;
     /**
      * Search the web and optionally scrape each result.
      * @param query Search query string.
@@ -2993,6 +3020,8 @@ interface CrawlErrorsResponse {
         url: string;
         code?: string;
         error: string;
+        /** Present when the page needs provider terms accepted first. */
+        requiresAction?: RequiresAction;
     }[];
     /**
      * URLs blocked by robots.txt
@@ -3528,4 +3557,4 @@ declare class Firecrawl extends FirecrawlClient {
     get v1(): FirecrawlApp;
 }
 
-export { type ActionOption, type ActiveCrawl, type ActiveCrawlsResponse, type AgentEffort, type AgentExchangeOptions, type AgentExchangeSummary, type AgentListItem, type AgentListOptions, type AgentListResponse, type AgentMode, type AgentOnTermsRequired, type AgentOptions$1 as AgentOptions, type AgentResponse, type AgentSkippedProvider, type AgentSnapshotResponse, type AgentStatusResponse, type AgentSuggestion, type AgentTermsGate, type AgentTermsRequiredAction, type AgentThread, type AgentThreadResponse, type AgentThreadRun, type AgentTraceActiveBrowserSession, type AgentTraceAgentFinishedEvent, type AgentTraceAgentIdentity, type AgentTraceAgentRole, type AgentTraceAgentStartedEvent, type AgentTraceArtifactChange, type AgentTraceArtifactUpdatedEvent, type AgentTraceBrowserSessionFinishedEvent, type AgentTraceBrowserSessionStartedEvent, type AgentTraceError, type AgentTraceErrorOccurredEvent, type AgentTraceEvent, type AgentTraceProgressReportedEvent, type AgentTraceReasoningSummaryEvent, type AgentTraceResponse, type AgentTraceRunCancelRequestedEvent, type AgentTraceRunFinishedEvent, type AgentTraceRunStartedEvent, type AgentTraceToolCallFinishedEvent, type AgentTraceToolCallStartedEvent, type AgentWebhookConfig, type AgentWebhookEvent, type AlexandriaCall, type AlexandriaOptions, type AlexandriaScrapeData, type AlexandriaScrapeError, type AlexandriaScrapeRequest, type AlexandriaScrapeResult, type AttributesFormat, type AuditMetadata$1 as AuditMetadata, type BatchScrapeJob, type BatchScrapeOptions, type BatchScrapeResponse$1 as BatchScrapeResponse, type BrandingProfile, type BrowserCreateResponse, type BrowserDeleteResponse, type BrowserExecuteResponse, type BrowserListResponse, type BrowserSession, type CategoryOption, type ChangeTrackingFormat, type ClickAction, type ConcurrencyCheck, type CrawlErrorsResponse$1 as CrawlErrorsResponse, type CrawlJob, type CrawlOptions, type CrawlResponse$1 as CrawlResponse, type CreateMonitorRequest, type CreditUsage, type CreditUsageHistoricalPeriod, type CreditUsageHistoricalResponse, type DeveloperSearchLicenseDisclosure, type DeveloperSearchOptions, type DeveloperSearchPassage, type DeveloperSearchRepoStatus, type DeveloperSearchResponse, type DeveloperSearchResult, type DeveloperSearchSourceStatus, type DeveloperSearchType, type DiscoveredTool, type Document, type DocumentMetadata, type EndpointFeedbackEndpoint, type EndpointFeedbackRequest, type ErrorDetails, type ExecuteJavascriptAction, type ExtractResponse$1 as ExtractResponse, type FeedbackMissingContent, type FeedbackRating, type FeedbackResponse, type FeedbackValuableSource, type FindToolsData, type FindToolsOptions, Firecrawl, FirecrawlApp as FirecrawlAppV1, FirecrawlClient, type FirecrawlClientInput, type FirecrawlClientOptions, type Format, type FormatOption, type FormatString, type GetMonitorCheckOptions, type GetPaperOptions, type GitHubScoreBreakdown, type GitHubSearchItem, type GitHubSearchResponse, type HighlightsFormat, type IdMap, type ImageParser, JobTimeoutError, type JsonFormat, type ListMonitorChecksOptions, type ListMonitorsOptions, type LocationConfig$1 as LocationConfig, type MapData, type MapOptions, type MenuAvailability, type MenuImage, type MenuItem, type MenuItemIdentifiers, type MenuMerchant, type MenuPrice, type MenuProfile, type MenuSection, type Monitor, type MonitorCheck, type MonitorCheckDetail, type MonitorCheckPage, type MonitorCrawlTarget, type MonitorCrawlTargetResult, type MonitorEmailNotification, type MonitorEmailRecipientSubscription, type MonitorJsonFieldDiff, type MonitorNotification, type MonitorPageDiff, type MonitorPageJudgment, type MonitorPageSnapshot, type MonitorSchedule, type MonitorScrapeTarget, type MonitorScrapeTargetResult, type MonitorSearchTarget, type MonitorSearchTargetResult, type MonitorSummary, type MonitorTarget, type MonitorTargetResult, type MonitorWebhookConfig, type PDFAction, type PDFParser, type PaginationConfig, type PaperMetadata, type PaperMetadataResponse, type PaperResult, type PaperSignals, type ParseFile, type ParseFileData, type ParseFormat, type ParseFormatOption, type ParseFormatString, type ParseOptions, type Passage, type PdfBlockConfidence, type PdfBlockItem, type PdfPage, type PdfPageBlocks, type PendingApproval, type PendingApprovalCall, type PendingCallsApproval, type PendingTermsApproval, type PressAction, type ProductAvailability, type ProductImage, type ProductPrice, type ProductProfile, type ProductSale, type ProductVariant, type QueryFormat, type QuestionFormat, type QueueStatusResponse$1 as QueueStatusResponse, type ReadPaperResponse, type RedactPIIEntity, type RedactPIIOptions, type RequiresAction, ResearchClient, type ScrapeAction, type ScrapeBrowserDeleteResponse, type ScrapeExecuteRequest, type ScrapeExecuteResponse, type ScrapeOptions, type ScreenshotAction, type ScreenshotFormat, type ScrollAction, SdkError, type SearchData, type SearchFeedbackRequest, type SearchGithubOptions, type SearchPapersOptions, type SearchPapersResponse, type SearchRequest, type SearchResultImages, type SearchResultNews, type SearchResultWeb, type SimilarPapersOptions, type SimilarPapersResponse, type ThreatProtectionOptions, type TokenUsage, type TokenUsageHistoricalPeriod, type TokenUsageHistoricalResponse, type UpdateMonitorRequest, type Viewport, type WaitAction, Watcher, type WatcherOptions, type WebhookConfig, type WriteAction, Firecrawl as default, parseRequiresAction };
+export { type ActionOption, type ActiveCrawl, type ActiveCrawlsResponse, type AgentEffort, type AgentExchangeOptions, type AgentExchangeSummary, type AgentListItem, type AgentListOptions, type AgentListResponse, type AgentMode, type AgentOnTermsRequired, type AgentOptions$1 as AgentOptions, type AgentResponse, type AgentSkippedProvider, type AgentSnapshotResponse, type AgentStatusResponse, type AgentSuggestion, type AgentTermsGate, type AgentTermsRequiredAction, type AgentThread, type AgentThreadResponse, type AgentThreadRun, type AgentTraceActiveBrowserSession, type AgentTraceAgentFinishedEvent, type AgentTraceAgentIdentity, type AgentTraceAgentRole, type AgentTraceAgentStartedEvent, type AgentTraceArtifactChange, type AgentTraceArtifactUpdatedEvent, type AgentTraceBrowserSessionFinishedEvent, type AgentTraceBrowserSessionStartedEvent, type AgentTraceError, type AgentTraceErrorOccurredEvent, type AgentTraceEvent, type AgentTraceProgressReportedEvent, type AgentTraceReasoningSummaryEvent, type AgentTraceResponse, type AgentTraceRunCancelRequestedEvent, type AgentTraceRunFinishedEvent, type AgentTraceRunStartedEvent, type AgentTraceToolCallFinishedEvent, type AgentTraceToolCallStartedEvent, type AgentWebhookConfig, type AgentWebhookEvent, type AlexandriaCall, type AlexandriaOptions, type AlexandriaScrapeData, type AlexandriaScrapeError, type AlexandriaScrapeRequest, type AlexandriaScrapeResult, type AttributesFormat, type AuditMetadata$1 as AuditMetadata, type BatchScrapeJob, type BatchScrapeOptions, type BatchScrapeResponse$1 as BatchScrapeResponse, type BrandingProfile, type BrowserCreateResponse, type BrowserDeleteResponse, type BrowserExecuteResponse, type BrowserListResponse, type BrowserSession, type CategoryOption, type ChangeTrackingFormat, type ClickAction, type ConcurrencyCheck, type CrawlErrorsResponse$1 as CrawlErrorsResponse, type CrawlJob, type CrawlOptions, type CrawlResponse$1 as CrawlResponse, type CreateMonitorRequest, type CreditUsage, type CreditUsageHistoricalPeriod, type CreditUsageHistoricalResponse, type DeveloperSearchLicenseDisclosure, type DeveloperSearchOptions, type DeveloperSearchPassage, type DeveloperSearchRepoStatus, type DeveloperSearchResponse, type DeveloperSearchResult, type DeveloperSearchSourceStatus, type DeveloperSearchType, type DiscoveredTool, type Document, type DocumentMetadata, type EndpointFeedbackEndpoint, type EndpointFeedbackRequest, type ErrorDetails, type ExecuteJavascriptAction, type ExtractResponse$1 as ExtractResponse, type FeedbackMissingContent, type FeedbackRating, type FeedbackResponse, type FeedbackValuableSource, type FindToolsData, type FindToolsOptions, Firecrawl, FirecrawlApp as FirecrawlAppV1, FirecrawlClient, type FirecrawlClientInput, type FirecrawlClientOptions, type Format, type FormatOption, type FormatString, type GetMonitorCheckOptions, type GetPaperOptions, type GitHubScoreBreakdown, type GitHubSearchItem, type GitHubSearchResponse, type HighlightsFormat, type IdMap, type ImageParser, JobTimeoutError, type JsonFormat, type ListMonitorChecksOptions, type ListMonitorsOptions, type LocationConfig$1 as LocationConfig, type MapData, type MapOptions, type MenuAvailability, type MenuImage, type MenuItem, type MenuItemIdentifiers, type MenuMerchant, type MenuPrice, type MenuProfile, type MenuSection, type Monitor, type MonitorCheck, type MonitorCheckDetail, type MonitorCheckPage, type MonitorCrawlTarget, type MonitorCrawlTargetResult, type MonitorEmailNotification, type MonitorEmailRecipientSubscription, type MonitorJsonFieldDiff, type MonitorNotification, type MonitorPageDiff, type MonitorPageJudgment, type MonitorPageSnapshot, type MonitorSchedule, type MonitorScrapeTarget, type MonitorScrapeTargetResult, type MonitorSearchTarget, type MonitorSearchTargetResult, type MonitorSummary, type MonitorTarget, type MonitorTargetResult, type MonitorWebhookConfig, type PDFAction, type PDFParser, type PaginationConfig, type PaperMetadata, type PaperMetadataResponse, type PaperResult, type PaperSignals, type ParseFile, type ParseFileData, type ParseFormat, type ParseFormatInfo, type ParseFormatOption, type ParseFormatString, type ParseOptions, type Passage, type PdfBlockConfidence, type PdfBlockItem, type PdfPage, type PdfPageBlocks, type PendingApproval, type PendingApprovalCall, type PendingCallsApproval, type PendingTermsApproval, type PressAction, type ProductAvailability, type ProductImage, type ProductPrice, type ProductProfile, type ProductSale, type ProductVariant, type QueryFormat, type QuestionFormat, type QueueStatusResponse$1 as QueueStatusResponse, type ReadPaperResponse, type RedactPIIEntity, type RedactPIIOptions, type RequiresAction, ResearchClient, type ScrapeAction, type ScrapeBrowserDeleteResponse, type ScrapeExecuteRequest, type ScrapeExecuteResponse, type ScrapeOptions, type ScreenshotAction, type ScreenshotFormat, type ScrollAction, SdkError, type SearchData, type SearchFeedbackRequest, type SearchGithubOptions, type SearchPapersOptions, type SearchPapersResponse, type SearchRequest, type SearchResultImages, type SearchResultNews, type SearchResultWeb, type SimilarPapersOptions, type SimilarPapersResponse, type ThreatProtectionOptions, type TokenUsage, type TokenUsageHistoricalPeriod, type TokenUsageHistoricalResponse, type UpdateMonitorRequest, type Viewport, type WaitAction, Watcher, type WatcherOptions, type WebhookConfig, type WriteAction, Firecrawl as default, parseRequiresAction };

package/dist/index.js

@@ -1,6 +1,6 @@
 import {
   require_package
-} from "./chunk-OCJTYRX5.js";
+} from "./chunk-XZ5XI5DK.js";
 
 // src/v2/utils/httpClient.ts
 import axios from "axios";
@@ -18,6 +18,21 @@ function getVersion() {
   }
 }
 
+// src/utils/apiOrigin.ts
+function pinToApiOrigin(apiUrl, url) {
+  const normalized = url.replace(/[\t\n\r]/g, "").replace(/^[\u0000-\u0020]+/, "");
+  if (!/^([a-z][a-z\d+\-.]*:|[\\/]{2})/i.test(normalized)) return url;
+  const base = URL.canParse(apiUrl) ? new URL(apiUrl) : null;
+  if (!base?.host) {
+    throw new Error(`apiUrl must be an absolute URL, got ${JSON.stringify(apiUrl)}`);
+  }
+  const target = new URL(normalized, base);
+  const pinned = new URL(base.origin);
+  pinned.pathname = target.pathname;
+  pinned.search = target.search;
+  return pinned.href;
+}
+
 // src/v2/utils/httpClient.ts
 var HttpClient = class {
   instance;
@@ -58,6 +73,7 @@ var HttpClient = class {
     config.headers = {
       ...config.headers || {}
     };
+    if (config.url) config.url = pinToApiOrigin(this.apiUrl, config.url);
     let lastError;
     for (let attempt = 0; attempt < this.maxRetries; attempt++) {
       try {
@@ -638,6 +654,16 @@ async function parse(http, file, options) {
     throw err;
   }
 }
+async function getParseFormats(http) {
+  try {
+    const res = await http.get("/v2/parse/formats");
+    if (res.status !== 200 || !res.data?.success) throwForBadResponse(res, "get parse formats");
+    return res.data.data?.formats ?? [];
+  } catch (err) {
+    if (err?.isAxiosError) return normalizeAxiosError(err, "get parse formats");
+    throw err;
+  }
+}
 
 // src/v2/methods/search.ts
 function prepareSearchPayload(req) {
@@ -2220,6 +2246,13 @@ var FirecrawlClient = class {
   async parse(file, options) {
     return parse(this.http, file, options);
   }
+  /**
+   * List the file formats the parse endpoint accepts on this deployment.
+   * @returns Formats with their kind, extensions, MIME types, and availability.
+   */
+  async getParseFormats() {
+    return getParseFormats(this.http);
+  }
   // Search
   /**
    * Search the web and optionally scrape each result.
@@ -2673,7 +2706,7 @@ var FirecrawlApp = class {
       if (typeof process !== "undefined" && process.env && process.env.npm_package_version) {
         return process.env.npm_package_version;
       }
-      const packageJson = await import("./package-ECWDBUY3.js");
+      const packageJson = await import("./package-JVVNUZPT.js");
       return packageJson.default.version;
     } catch (error) {
       const isTest = typeof process !== "undefined" && (process.env.JEST_WORKER_ID != null || false);
@@ -2888,7 +2921,7 @@ var FirecrawlApp = class {
       throw new FirecrawlError("No crawl ID provided", 400);
     }
     const headers = this.prepareHeaders();
-    const targetURL = new URL(nextURL ?? `${this.apiUrl}/v1/crawl/${id}`);
+    const targetURL = new URL(nextURL ?? `${this.apiUrl}/v1/crawl/${id}`, this.apiUrl);
     if (skip !== void 0) {
       targetURL.searchParams.set("skip", skip.toString());
     }
@@ -3133,7 +3166,7 @@ var FirecrawlApp = class {
       throw new FirecrawlError("No batch scrape ID provided", 400);
     }
     const headers = this.prepareHeaders();
-    const targetURL = new URL(nextURL ?? `${this.apiUrl}/v1/batch/scrape/${id}`);
+    const targetURL = new URL(nextURL ?? `${this.apiUrl}/v1/batch/scrape/${id}`, this.apiUrl);
     if (skip !== void 0) {
       targetURL.searchParams.set("skip", skip.toString());
     }
@@ -3345,7 +3378,7 @@ var FirecrawlApp = class {
    */
   async getRequest(url, headers) {
     try {
-      return await axios2.get(url, { headers });
+      return await axios2.get(pinToApiOrigin(this.apiUrl, url), { headers });
     } catch (error) {
       if (error instanceof AxiosError && error.response) {
         return error.response;

package/dist/{package-ECWDBUY3.js → package-JVVNUZPT.js}

@@ -1,4 +1,4 @@
 import {
   require_package
-} from "./chunk-OCJTYRX5.js";
+} from "./chunk-XZ5XI5DK.js";
 export default require_package();

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "firecrawl",
-  "version": "4.42.0",
+  "version": "4.42.2",
   "description": "JavaScript SDK for the Firecrawl API: web scraping, crawling, web search, and scientific literature search over a research paper index of PubMed, bioRxiv, medRxiv and arXiv abstracts",
   "main": "dist/index.js",
   "types": "dist/index.d.ts",
@@ -19,7 +19,7 @@
   "author": "Mendable.ai",
   "license": "MIT",
   "dependencies": {
-    "axios": "1.18.0",
+    "axios": "1.20.0",
     "typescript-event-target": "^1.1.1",
     "zod": "^3.23.8",
     "zod-to-json-schema": "^3.23.0"

package/pnpm-workspace.yaml

@@ -7,10 +7,17 @@ overrides:
   "rollup@<4.59.0": ">=4.59.0"
   "picomatch@<4.0.4": ">=4.0.4"
   handlebars: ">=4.7.9"
-  "brace-expansion@>=5.0.0 <5.0.9": "5.0.9"
+  # GHSA-q2hr-2g5m-vwhr / GHSA-qhr7-859c-m2p7 / GHSA-6j4f-fj2g-mc7p:
+  # pin the 5.x line only. Exact 5.0.12 cannot resolve to 6.x.
+  "brace-expansion@>=5.0.0 <5.0.12": "5.0.12"
   "js-yaml@>=3.0.0 <3.15.2": "3.15.2"
   "js-yaml@>=4.0.0 <4.3.2": "4.3.2"
-  "axios@>=1.0.0 <1.16.0": "1.16.1"
+  # GHSA-vh66-26gq-q6x8 / GHSA-9fr6-4gfg-395g / GHSA-c29m-xwm3-cm6r /
+  # GHSA-mghh-pgcx-3jjj / GHSA-x97p-jq2g-jp4f / GHSA-3pq3-5fj3-cg6v /
+  # GHSA-542g-h47m-68v8 / GHSA-j8rh-479h-cp32 / GHSA-4hqw-qxg8-jxx2 /
+  # GHSA-m8m8-qj5v-23w3 / GHSA-44g4-m2mj-wpvx / GHSA-r4gj-5m52-g5wh:
+  # pin the 1.x line only. Exact 1.20.0 cannot resolve to 2.x.
+  "axios@>=1.0.0 <1.20.0": "1.20.0"
   "follow-redirects@<1.16.0": ">=1.16.0 <2.0.0"
   "form-data@>=4.0.0 <4.0.6": "4.0.6"
   "@babel/core@>=7.0.0 <7.29.6": "7.29.7"

package/src/__tests__/e2e/v2/parse.test.ts

@@ -3,6 +3,7 @@ import { config } from "dotenv";
 import { getIdentity, getApiUrl } from "./utils/idmux";
 import { testTimeoutMs, withRateLimitRetry } from "./utils/rateLimit";
 import { describe, test, expect, beforeAll } from "@jest/globals";
+import { SdkError } from "../../../v2/types";
 
 config();
 
@@ -65,4 +66,29 @@ describe("v2.parse e2e", () => {
     },
     testTimeoutMs(60_000),
   );
+
+  test(
+    "lists supported parse formats",
+    async () => {
+      if (!client) throw new Error();
+
+      let formats;
+      try {
+        formats = await client.getParseFormats();
+      } catch (err) {
+        if (err instanceof SdkError && err.status === 404) {
+          console.warn("Skipping: /v2/parse/formats is not deployed on this API");
+          return;
+        }
+        throw err;
+      }
+
+      expect(formats.length).toBeGreaterThan(0);
+      const pdf = formats.find(f => f.format === "pdf");
+      expect(pdf?.kind).toBe("document");
+      expect(pdf?.extensions).toContain(".pdf");
+      expect(pdf?.mimeTypes).toContain("application/pdf");
+    },
+    testTimeoutMs(60_000),
+  );
 });

package/src/__tests__/unit/v1/next-url-pinning.test.ts

@@ -0,0 +1,81 @@
+import { describe, test, expect, jest, afterEach } from "@jest/globals";
+import axios from "axios";
+import FirecrawlApp from "../../../v1";
+
+const API_URL = "https://api.firecrawl.dev";
+const API_KEY = "fc-test";
+
+const origins: Array<[string, string]> = [
+  ["same origin", API_URL],
+  ["cross host", "https://evil.example"],
+  ["protocol-relative", "//evil.example"],
+  ["different port", "https://api.firecrawl.dev:8443"],
+  ["different scheme", "http://api.firecrawl.dev"],
+  ["whitespace-prefixed", "\n https://evil.example"],
+];
+
+function mockGet(respond: (url: URL) => unknown) {
+  const sent: Array<{ url: string; authorization: unknown }> = [];
+  jest.spyOn(axios, "get").mockImplementation(async (url: string, config?: any) => {
+    sent.push({ url, authorization: config?.headers?.Authorization });
+    return { status: 200, data: respond(new URL(url)) };
+  });
+  return sent;
+}
+
+function expectPinned(sent: Array<{ url: string; authorization: unknown }>, expectedUrls: string[]) {
+  expect(sent.map(req => req.url)).toEqual(expectedUrls);
+  for (const req of sent) {
+    expect(new URL(req.url).origin).toBe(API_URL);
+    expect(req.url).not.toContain("evil.example");
+    expect(req.authorization).toBe(`Bearer ${API_KEY}`);
+  }
+}
+
+const completedPage = (next: string) => ({
+  success: true,
+  status: "completed",
+  total: 2,
+  completed: 2,
+  expiresAt: "2030-01-01T00:00:00Z",
+  data: [{ markdown: "a" }],
+  next,
+});
+const lastPage = () => ({ success: true, status: "completed", data: [{ markdown: "b" }] });
+
+describe("v1 pins next URLs to the api_url origin", () => {
+  afterEach(() => {
+    jest.restoreAllMocks();
+  });
+
+  describe.each(origins)("%s next", (_label, origin) => {
+    const app = new FirecrawlApp({ apiKey: API_KEY, apiUrl: API_URL });
+
+    test.each([
+      ["checkCrawlStatus", "/v1/crawl/abc"],
+      ["checkBatchScrapeStatus", "/v1/batch/scrape/abc"],
+    ] as const)("%s getAllData follows server next", async (method, path) => {
+      const sent = mockGet(url => (url.searchParams.has("skip") ? lastPage() : completedPage(`${origin}${path}?skip=10`)));
+      const res: any = await app[method]("abc", true);
+      expect(res.data).toHaveLength(2);
+      expectPinned(sent, [`${API_URL}${path}`, `${API_URL}${path}?skip=10`]);
+    });
+
+    test.each([
+      ["checkCrawlStatus", "/v1/crawl/abc"],
+      ["checkBatchScrapeStatus", "/v1/batch/scrape/abc"],
+    ] as const)("%s caller-supplied nextURL", async (method, path) => {
+      const sent = mockGet(lastPage);
+      await app[method]("abc", false, `${origin}${path}?skip=10`);
+      expectPinned(sent, [`${API_URL}${path}?skip=10`]);
+    });
+
+    test("monitorJobStatus follows server next", async () => {
+      const path = "/v1/crawl/abc";
+      const sent = mockGet(url => (url.searchParams.has("skip") ? lastPage() : completedPage(`${origin}${path}?skip=10`)));
+      const res: any = await app.monitorJobStatus("abc", app.prepareHeaders(), 0);
+      expect(res.data).toHaveLength(2);
+      expectPinned(sent, [`${API_URL}${path}`, `${API_URL}${path}?skip=10`]);
+    });
+  });
+});

package/src/__tests__/unit/v2/next-url-pinning.test.ts

@@ -0,0 +1,104 @@
+import { describe, test, expect } from "@jest/globals";
+import axios, { type AxiosAdapter } from "axios";
+import { FirecrawlClient } from "../../../v2/client";
+import { pinToApiOrigin } from "../../../utils/apiOrigin";
+
+const API_URL = "https://api.firecrawl.dev";
+const API_KEY = "fc-test";
+
+const origins: Array<[string, string]> = [
+  ["same origin", API_URL],
+  ["cross host", "https://evil.example"],
+  ["protocol-relative", "//evil.example"],
+  ["different port", "https://api.firecrawl.dev:8443"],
+  ["different scheme", "http://api.firecrawl.dev"],
+  ["whitespace-prefixed", "\n https://evil.example"],
+];
+
+function makeClient(firstPage: (next: string) => unknown, lastPage: unknown, next: string) {
+  const client = new FirecrawlClient({ apiKey: API_KEY, apiUrl: API_URL });
+  const sent: Array<{ url: string; authorization: unknown }> = [];
+  const adapter: AxiosAdapter = async config => {
+    const url = axios.getUri(config);
+    sent.push({ url, authorization: config.headers.Authorization });
+    const data = new URL(url).searchParams.has("skip") ? lastPage : firstPage(next);
+    return { data, status: 200, statusText: "OK", headers: {}, config };
+  };
+  (client as any).http.instance.defaults.adapter = adapter;
+  return { client, sent };
+}
+
+function expectPinned(sent: Array<{ url: string; authorization: unknown }>, path: string) {
+  expect(sent).toHaveLength(2);
+  expect(sent[1]!.url).toBe(`${API_URL}${path}?skip=10`);
+  for (const req of sent) {
+    expect(new URL(req.url).origin).toBe(API_URL);
+    expect(req.url).not.toContain("evil.example");
+    expect(req.authorization).toBe(`Bearer ${API_KEY}`);
+  }
+}
+
+describe("v2 pagination pins next URLs to the api_url origin", () => {
+  describe.each(origins)("%s next", (_label, origin) => {
+    test("crawl status", async () => {
+      const path = "/v2/crawl/abc";
+      const { client, sent } = makeClient(
+        next => ({ success: true, status: "completed", completed: 2, total: 2, next, data: [{ markdown: "a" }] }),
+        { success: true, next: null, data: [{ markdown: "b" }] },
+        `${origin}${path}?skip=10`,
+      );
+      const job = await client.getCrawlStatus("abc");
+      expect(job.data).toHaveLength(2);
+      expectPinned(sent, path);
+    });
+
+    test("batch scrape status", async () => {
+      const path = "/v2/batch/scrape/abc";
+      const { client, sent } = makeClient(
+        next => ({ success: true, status: "completed", completed: 2, total: 2, next, data: [{ markdown: "a" }] }),
+        { success: true, next: null, data: [{ markdown: "b" }] },
+        `${origin}${path}?skip=10`,
+      );
+      const job = await client.getBatchScrapeStatus("abc");
+      expect(job.data).toHaveLength(2);
+      expectPinned(sent, path);
+    });
+
+    test("monitor check", async () => {
+      const path = "/v2/monitor/m1/checks/c1";
+      const { client, sent } = makeClient(
+        next => ({ success: true, data: { id: "c1", pages: [{ url: "https://a.example" }], next } }),
+        { success: true, data: { pages: [{ url: "https://b.example" }], next: null } },
+        `${origin}${path}?skip=10`,
+      );
+      const check = await client.getMonitorCheck("m1", "c1");
+      expect(check.pages).toHaveLength(2);
+      expectPinned(sent, path);
+    });
+  });
+});
+
+describe("pinToApiOrigin", () => {
+  test("rewrites absolute URLs onto the api origin, keeping path and query and dropping the fragment", () => {
+    expect(pinToApiOrigin(API_URL, "https://evil.example/v2/crawl/abc?skip=10#frag")).toBe(
+      `${API_URL}/v2/crawl/abc?skip=10`,
+    );
+    expect(pinToApiOrigin("http://localhost:3002", "https://evil.example:444/v2/crawl/abc?skip=10")).toBe(
+      "http://localhost:3002/v2/crawl/abc?skip=10",
+    );
+    expect(pinToApiOrigin(API_URL, "\\\\evil.example/v2/crawl/abc")).toBe(`${API_URL}/v2/crawl/abc`);
+    for (const url of ["\nhttps://evil.example/v2/crawl/abc", "h\tttps://evil.example/v2/crawl/abc", "\u0000 //evil.example/v2/crawl/abc", "/\\evil.example/v2/crawl/abc"]) {
+      expect(pinToApiOrigin(API_URL, url)).toBe(`${API_URL}/v2/crawl/abc`);
+    }
+  });
+
+  test("leaves relative URLs unchanged", () => {
+    expect(pinToApiOrigin(API_URL, "/v2/crawl/abc?skip=10")).toBe("/v2/crawl/abc?skip=10");
+  });
+
+  test("throws when apiUrl is not absolute and the URL needs pinning", () => {
+    expect(() => pinToApiOrigin("api.firecrawl.dev", "https://evil.example/v2/crawl/abc")).toThrow(
+      "apiUrl must be an absolute URL",
+    );
+  });
+});

package/src/__tests__/unit/v2/parse-formats.unit.test.ts

@@ -0,0 +1,84 @@
+import { describe, test, expect } from "@jest/globals";
+import axios, { type AxiosAdapter } from "axios";
+import { Firecrawl } from "../../../index";
+import { SdkError } from "../../../v2/types";
+
+const API_URL = "https://api.firecrawl.dev";
+const API_KEY = "fc-test";
+
+function makeClient(status: number, data: unknown) {
+  const client = new Firecrawl({ apiKey: API_KEY, apiUrl: API_URL });
+  const sent: Array<{ url: string; method?: string; authorization: unknown }> = [];
+  const adapter: AxiosAdapter = async config => {
+    sent.push({ url: axios.getUri(config), method: config.method, authorization: config.headers.Authorization });
+    const response = { data, status, statusText: "", headers: {}, config };
+    if (status >= 400) {
+      throw new axios.AxiosError(`Request failed with status code ${status}`, "ERR_BAD_RESPONSE", config, null, response);
+    }
+    return response;
+  };
+  (client as any).http.instance.defaults.adapter = adapter;
+  return { client, sent };
+}
+
+describe("v2.getParseFormats unit", () => {
+  test("GETs /v2/parse/formats with auth and unwraps data.formats", async () => {
+    const { client, sent } = makeClient(200, {
+      success: true,
+      data: {
+        formats: [
+          { format: "pdf", kind: "document", extensions: [".pdf"], mimeTypes: ["application/pdf"], available: true },
+          { format: "png", kind: "image", extensions: [".png"], mimeTypes: ["image/png"], available: false },
+        ],
+      },
+    });
+
+    const formats = await client.getParseFormats();
+
+    expect(sent).toEqual([{ url: `${API_URL}/v2/parse/formats`, method: "get", authorization: `Bearer ${API_KEY}` }]);
+    expect(formats).toHaveLength(2);
+    expect(formats[0]).toEqual({ format: "pdf", kind: "document", extensions: [".pdf"], mimeTypes: ["application/pdf"], available: true });
+    expect(formats[1]!.kind).toBe("image");
+    expect(formats[1]!.mimeTypes).toEqual(["image/png"]);
+    expect(formats[1]!.available).toBe(false);
+  });
+
+  test("tolerates unknown kinds and extra fields", async () => {
+    const { client } = makeClient(200, {
+      success: true,
+      data: {
+        formats: [
+          { format: "mp3", kind: "audio", extensions: [".mp3"], mimeTypes: ["audio/mpeg"], available: true, maxSizeMb: 50 },
+        ],
+      },
+    });
+
+    const [format] = await client.getParseFormats();
+
+    expect(format!.format).toBe("mp3");
+    expect(format!.kind).toBe("audio");
+    expect(format!.mimeTypes).toEqual(["audio/mpeg"]);
+  });
+
+  test("returns an empty list when formats is missing", async () => {
+    const { client } = makeClient(200, { success: true, data: {} });
+
+    await expect(client.getParseFormats()).resolves.toEqual([]);
+  });
+
+  test.each([401, 500])("throws SdkError on %i", async status => {
+    const { client } = makeClient(status, { success: false, error: "Unauthorized" });
+
+    const err = await client.getParseFormats().catch(e => e);
+
+    expect(err).toBeInstanceOf(SdkError);
+    expect(err.status).toBe(status);
+    expect(err.message).toBe("Unauthorized");
+  });
+
+  test("throws SdkError when success is false", async () => {
+    const { client } = makeClient(200, { success: false, error: "nope" });
+
+    await expect(client.getParseFormats()).rejects.toBeInstanceOf(SdkError);
+  });
+});

package/src/utils/apiOrigin.ts

@@ -0,0 +1,19 @@
+/**
+ * Rewrite an absolute or protocol-relative URL onto apiUrl's scheme, host and
+ * port (keeping path and query, dropping any fragment) so credentials never
+ * leave the configured API origin. Relative URLs are returned unchanged.
+ * Throws if apiUrl is not an absolute URL.
+ */
+export function pinToApiOrigin(apiUrl: string, url: string): string {
+  const normalized = url.replace(/[\t\n\r]/g, "").replace(/^[\u0000-\u0020]+/, "");
+  if (!/^([a-z][a-z\d+\-.]*:|[\\/]{2})/i.test(normalized)) return url;
+  const base = URL.canParse(apiUrl) ? new URL(apiUrl) : null;
+  if (!base?.host) {
+    throw new Error(`apiUrl must be an absolute URL, got ${JSON.stringify(apiUrl)}`);
+  }
+  const target = new URL(normalized, base);
+  const pinned = new URL(base.origin);
+  pinned.pathname = target.pathname;
+  pinned.search = target.search;
+  return pinned.href;
+}

package/src/v1/index.ts

@@ -2,6 +2,8 @@ import axios, { type AxiosResponse, type AxiosRequestHeaders, AxiosError } from
 import * as zt from "zod";
 import { zodSchemaToJsonSchema } from "../utils/zodSchemaToJson";
 import { TypedEventTarget } from "typescript-event-target";
+import type { RequiresAction } from "../v2/types";
+import { pinToApiOrigin } from "../utils/apiOrigin";
 
 /**
  * Configuration interface for FirecrawlApp.
@@ -439,6 +441,8 @@ export interface CrawlErrorsResponse {
     url: string,
     code?: string,
     error: string,
+    /** Present when the page needs provider terms accepted first. */
+    requiresAction?: RequiresAction,
   }[];
 
   /**
@@ -919,7 +923,7 @@ export default class FirecrawlApp {
     }
 
     const headers: AxiosRequestHeaders = this.prepareHeaders();
-    const targetURL = new URL(nextURL ?? `${this.apiUrl}/v1/crawl/${id}`);
+    const targetURL = new URL(nextURL ?? `${this.apiUrl}/v1/crawl/${id}`, this.apiUrl);
     if (skip !== undefined) {
       targetURL.searchParams.set("skip", skip.toString());
     }
@@ -1207,7 +1211,7 @@ export default class FirecrawlApp {
     }
 
     const headers: AxiosRequestHeaders = this.prepareHeaders();
-    const targetURL = new URL(nextURL ?? `${this.apiUrl}/v1/batch/scrape/${id}`);
+    const targetURL = new URL(nextURL ?? `${this.apiUrl}/v1/batch/scrape/${id}`, this.apiUrl);
     if (skip !== undefined) {
       targetURL.searchParams.set("skip", skip.toString());
     }
@@ -1448,7 +1452,7 @@ export default class FirecrawlApp {
     headers: AxiosRequestHeaders
   ): Promise<AxiosResponse> {
     try {
-      return await axios.get(url, { headers });
+      return await axios.get(pinToApiOrigin(this.apiUrl, url), { headers });
     } catch (error) {
       if (error instanceof AxiosError && error.response) {
         return error.response as AxiosResponse;

package/src/v2/client.ts

@@ -4,7 +4,7 @@ import {
   interact as interactMethod,
   stopInteraction as stopInteractionMethod,
 } from "./methods/scrape";
-import { parse as parseMethod } from "./methods/parse";
+import { parse as parseMethod, getParseFormats } from "./methods/parse";
 import { search } from "./methods/search";
 import { scrapeAlexandria, findTools } from "./methods/tools";
 import { developerSearch as developerSearchMethod } from "./methods/developer";
@@ -53,6 +53,7 @@ import type {
   FindToolsData,
   AlexandriaScrapeData,
   ParseFile,
+  ParseFormatInfo,
   ParseOptions,
   ScrapeOptions,
   SearchData,
@@ -273,6 +274,14 @@ export class FirecrawlClient {
     return parseMethod(this.http, file, options);
   }
 
+  /**
+   * List the file formats the parse endpoint accepts on this deployment.
+   * @returns Formats with their kind, extensions, MIME types, and availability.
+   */
+  async getParseFormats(): Promise<ParseFormatInfo[]> {
+    return getParseFormats(this.http);
+  }
+
   // Search
   /**
    * Search the web and optionally scrape each result.

package/src/v2/methods/agent.ts

@@ -12,6 +12,7 @@ function prepareAgentPayload(args: {
   origin?: string;
   maxCredits?: number;
   strictConstrainToURLs?: boolean;
+  /** Defaults to spark-2. "spark-1-pro" and "spark-1-mini" are deprecated and run spark-2. */
   model?: "spark-1-pro" | "spark-1-mini" | "spark-2";
   effort?: AgentEffort;
   webhook?: string | AgentWebhookConfig;

package/src/v2/methods/batch.ts

@@ -152,7 +152,7 @@ export async function getBatchScrapeErrors(
   try {
     const res = await http.get<{
       success?: boolean;
-      data?: { errors: Array<Record<string, string>>; robotsBlocked: string[] };
+      data?: CrawlErrorsResponse;
     }>(`/v2/batch/scrape/${jobId}/errors`);
     if (res.status !== 200) throwForBadResponse(res, "get batch scrape errors");
     const payload = res.data?.data ?? (res.data as any);

package/src/v2/methods/crawl.ts

@@ -162,7 +162,7 @@ export async function crawl(http: HttpClient, request: CrawlRequest, pollInterva
 
 export async function getCrawlErrors(http: HttpClient, crawlId: string): Promise<CrawlErrorsResponse> {
   try {
-    const res = await http.get<{ success?: boolean; data?: { errors: Array<Record<string, string>>; robotsBlocked: string[] } }>(`/v2/crawl/${crawlId}/errors`);
+    const res = await http.get<{ success?: boolean; data?: CrawlErrorsResponse }>(`/v2/crawl/${crawlId}/errors`);
     if (res.status !== 200) throwForBadResponse(res, "get crawl errors");
     const payload = res.data?.data ?? (res.data as any);
     return { errors: payload.errors || [], robotsBlocked: payload.robotsBlocked || [] };

package/src/v2/methods/parse.ts

@@ -1,4 +1,4 @@
-import { type Document, type ParseFile, type ParseOptions } from "../types";
+import { type Document, type ParseFile, type ParseFormatInfo, type ParseOptions } from "../types";
 import { HttpClient } from "../utils/httpClient";
 import { agentHintMetadata } from "../utils/agentHints";
 import { ensureValidParseOptions } from "../utils/validation";
@@ -95,3 +95,14 @@ export async function parse(
     throw err;
   }
 }
+
+export async function getParseFormats(http: HttpClient): Promise<ParseFormatInfo[]> {
+  try {
+    const res = await http.get<{ success: boolean; data?: { formats: ParseFormatInfo[] } }>("/v2/parse/formats");
+    if (res.status !== 200 || !res.data?.success) throwForBadResponse(res, "get parse formats");
+    return res.data.data?.formats ?? [];
+  } catch (err: any) {
+    if (err?.isAxiosError) return normalizeAxiosError(err, "get parse formats");
+    throw err;
+  }
+}

package/src/v2/types.ts

@@ -366,6 +366,18 @@ export type ParseOptions = Omit<
   proxy?: "basic" | "auto";
 };
 
+/** A file format accepted by the parse endpoint. */
+export interface ParseFormatInfo {
+  /** Format identifier, e.g. "pdf". */
+  format: string;
+  kind: "document" | "image" | (string & {});
+  /** File extensions including the leading dot, e.g. ".pdf". */
+  extensions: string[];
+  mimeTypes: string[];
+  /** False when the format is recognized but not enabled on this deployment. */
+  available: boolean;
+}
+
 export interface WebhookConfig {
   url: string;
   headers?: Record<string, string>;
@@ -1627,6 +1639,11 @@ export interface AgentStatusResponse {
   status: "processing" | "completed" | "failed";
   error?: string;
   data?: unknown;
+  /** Best-effort result on a failed run; never a completed `data` value. */
+  partial?: unknown;
+  /** Schema validity of `partial`, when a schema was supplied. */
+  partialSchemaValid?: boolean;
+  stopReason?: "credit_limit_reached";
   /**
    * Server-provided model name. Widened past the request-side union on
    * purpose: new models ship without an SDK release, so pinning this to known
@@ -1672,6 +1689,9 @@ export interface AgentThreadRun {
   message: string | null;
   /** Only present when the request asked for includeData. */
   data?: unknown;
+  partial?: unknown;
+  partialSchemaValid?: boolean;
+  stopReason?: "credit_limit_reached";
   suggestions?: AgentSuggestion[] | null;
   pendingApproval?: PendingApproval | null;
   exchange?: AgentExchangeSummary | null;
@@ -1955,6 +1975,8 @@ export interface CrawlErrorsResponse {
     url: string;
     code?: string;
     error: string;
+    /** Present when the page needs provider terms accepted first. */
+    requiresAction?: RequiresAction;
   }[];
   robotsBlocked: string[];
 }

package/src/v2/utils/httpClient.ts

@@ -4,6 +4,7 @@ import axios, {
   type AxiosResponse,
 } from "axios";
 import { getVersion } from "./getVersion";
+import { pinToApiOrigin } from "../../utils/apiOrigin";
 
 export interface HttpClientOptions {
   apiKey: string;
@@ -64,6 +65,7 @@ export class HttpClient {
     config.headers = {
       ...(config.headers || {}),
     };
+    if (config.url) config.url = pinToApiOrigin(this.apiUrl, config.url);
 
     let lastError: any;
     for (let attempt = 0; attempt < this.maxRetries; attempt++) {