@mindstudio-ai/agent 0.1.89 → 0.1.90

package/dist/index.d.ts

@@ -3211,6 +3211,8 @@ interface ExtractAudioFromVideoStepOutput {
 interface ExtractTextStepInput {
     /** URL or array of URLs to extract text from. Accepts a single URL, comma-separated list, or JSON array */
     url: string | string[];
+    /** Optional extraction model id (a `document_extraction` model, e.g. `mistral-ocr-latest`, `llamaparse`, `google-document-ai`). Defaults to the platform default when omitted. */
+    model?: string;
 }
 interface ExtractTextStepOutput {
     /** Extracted text content. A single string for one URL, or an array for multiple URLs */
@@ -7269,6 +7271,7 @@ interface StepMethods {
      * - Handles both text-layer PDFs and image-based/scanned PDFs (e.g. Figma/Canva exports, scanned documents). Image-based PDFs are processed with OCR automatically — there is no need to convert PDF pages to images first.
      * - Accepts a single URL, a comma-separated list of URLs, or a JSON array of URLs.
      * - Files are rehosted on the MindStudio CDN before extraction.
+     * - Optionally set `model` to a specific document-extraction model (`mistral-ocr-latest`, `llamaparse`, `google-document-ai`); omit to use the platform default.
      * - Maximum file size is 50MB per URL.
      *
      * @example

package/dist/index.js

@@ -107,8 +107,16 @@ var WS_OPEN_TIMEOUT_MS = 5e3;
 var WS_OPEN = 1;
 var WS_MAX_FRAME_CHARS = 9e5;
 var DbWsTransportError = class extends Error {
+  constructor(message, sent = false) {
+    super(message);
+    this.sent = sent;
+  }
   name = "DbWsTransportError";
 };
+function isReadOnlySql(sql) {
+  const head = sql.trimStart().slice(0, 8).toUpperCase();
+  return head.startsWith("SELECT") || head.startsWith("PRAGMA") || head.startsWith("EXPLAIN");
+}
 var socket = null;
 var opening = null;
 var idCounter = 0;
@@ -136,7 +144,7 @@ function getOrOpen(url) {
       opening = null;
       for (const [, p] of pending) {
         clearTimeout(p.timer);
-        p.reject(err);
+        p.reject(new DbWsTransportError(err.message, true));
       }
       pending.clear();
       reject(err);
@@ -206,7 +214,7 @@ async function executeDbBatchOverWs(url, token, databaseId, queries) {
   return await new Promise((resolve, reject) => {
     const timer = setTimeout(() => {
       pending.delete(id);
-      reject(new DbWsTransportError("ws execute timed out"));
+      reject(new DbWsTransportError("ws execute timed out", true));
     }, WS_REQUEST_TIMEOUT_MS);
     pending.set(id, { resolve, reject, timer });
     try {
@@ -215,7 +223,10 @@ async function executeDbBatchOverWs(url, token, databaseId, queries) {
       clearTimeout(timer);
       pending.delete(id);
       reject(
-        new DbWsTransportError(`ws send failed: ${err?.message || "unknown"}`)
+        new DbWsTransportError(
+          `ws send failed: ${err?.message || "unknown"}`,
+          false
+        )
       );
     }
   });
@@ -2509,8 +2520,8 @@ var stepMetadata = {
   "extractText": {
     stepType: "extractText",
     description: "Download a file from a URL and extract its text content. Supports PDFs (including scanned/image-based PDFs via OCR), plain text files, and other document formats.",
-    usageNotes: "- Best suited for PDFs and raw text/document files. For web pages, use the scrapeUrl step instead.\n- Handles both text-layer PDFs and image-based/scanned PDFs (e.g. Figma/Canva exports, scanned documents). Image-based PDFs are processed with OCR automatically \u2014 there is no need to convert PDF pages to images first.\n- Accepts a single URL, a comma-separated list of URLs, or a JSON array of URLs.\n- Files are rehosted on the MindStudio CDN before extraction.\n- Maximum file size is 50MB per URL.",
-    inputSchema: { "type": "object", "properties": { "url": { "anyOf": [{ "type": "string" }, { "type": "array", "items": { "type": "string" } }] } }, "required": ["url"] },
+    usageNotes: "- Best suited for PDFs and raw text/document files. For web pages, use the scrapeUrl step instead.\n- Handles both text-layer PDFs and image-based/scanned PDFs (e.g. Figma/Canva exports, scanned documents). Image-based PDFs are processed with OCR automatically \u2014 there is no need to convert PDF pages to images first.\n- Accepts a single URL, a comma-separated list of URLs, or a JSON array of URLs.\n- Files are rehosted on the MindStudio CDN before extraction.\n- Optionally set `model` to a specific document-extraction model (`mistral-ocr-latest`, `llamaparse`, `google-document-ai`); omit to use the platform default.\n- Maximum file size is 50MB per URL.",
+    inputSchema: { "type": "object", "properties": { "url": { "anyOf": [{ "type": "string" }, { "type": "array", "items": { "type": "string" } }] }, "model": { "type": "string", "description": "Optional extraction model id (a `document_extraction` model, e.g. `mistral-ocr-latest`, `llamaparse`, `google-document-ai`). Defaults to the platform default when omitted." } }, "required": ["url"] },
     outputSchema: { "type": "object", "properties": { "text": { "anyOf": [{ "type": "string" }, { "type": "array", "items": { "type": "string" } }] } }, "required": ["text"] }
   },
   "fetchDataSourceDocument": {
@@ -4353,7 +4364,9 @@ var MindStudioAgent = class {
       ...options?.appId != null && { appId: options.appId },
       ...threadId != null && { threadId },
       ...this._currentStreamId != null && { streamId: this._currentStreamId },
-      ...this._requestSource != null && { requestSource: this._requestSource },
+      ...this._requestSource != null && {
+        requestSource: this._requestSource
+      },
       ...assetStoreBody(options?.store)
     });
     const remaining = headers.get("x-ratelimit-remaining");
@@ -4438,7 +4451,9 @@ var MindStudioAgent = class {
       ...options.appId != null && { appId: options.appId },
       ...threadId != null && { threadId },
       ...this._currentStreamId != null && { streamId: this._currentStreamId },
-      ...this._requestSource != null && { requestSource: this._requestSource },
+      ...this._requestSource != null && {
+        requestSource: this._requestSource
+      },
       ...assetStoreBody(options.store)
     };
     await this._httpConfig.rateLimiter.acquire();
@@ -4626,7 +4641,9 @@ var MindStudioAgent = class {
       })),
       ...options?.appId != null && { appId: options.appId },
       ...threadId != null && { threadId },
-      ...this._requestSource != null && { requestSource: this._requestSource },
+      ...this._requestSource != null && {
+        requestSource: this._requestSource
+      },
       ...assetStoreBody(options?.store)
     });
     const pollUrl = `${this._currentHttpConfig.baseUrl}/developer/v2/steps/execute-batch-async/poll/${asyncData.batchToken}`;
@@ -5242,6 +5259,13 @@ var MindStudioAgent = class {
         if (!(err instanceof DbWsTransportError)) {
           throw err;
         }
+        if (err.sent && queries.some((q) => !isReadOnlySql(q.sql))) {
+          throw new MindStudioError(
+            "[db] Connection was interrupted after this query was sent; because it contains a write, it was not automatically retried (the write may or may not have been applied). Verify the current state before re-running it.",
+            "db_transport_interrupted",
+            503
+          );
+        }
         console.warn(
           `[mindstudio] db: WebSocket transport unavailable (${err.message}); using HTTP for this query.`
         );
@@ -5632,7 +5656,10 @@ var MindStudioAgent = class {
       content.byteOffset,
       content.byteOffset + content.byteLength
     );
-    const fileBlob = new Blob([buf], options.type ? { type: options.type } : void 0);
+    const fileBlob = new Blob(
+      [buf],
+      options.type ? { type: options.type } : void 0
+    );
     form.append("file", fileBlob, filename);
     const res = await fetch(data.url, { method: "POST", body: form });
     if (!res.ok) {