libretto 0.3.0 → 0.3.2

package/dist/index.cjs

@@ -0,0 +1,144 @@
+"use strict";
+var __create = Object.create;
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __getProtoOf = Object.getPrototypeOf;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toESM = (mod, isNodeMode, target) => (target = mod != null ? __create(__getProtoOf(mod)) : {}, __copyProps(
+  // If the importer is in node compatibility mode or this is not an ESM
+  // file that has been converted to a CommonJS file using a Babel-
+  // compatible transform (i.e. "__esModule" has not been set), then set
+  // "default" to the CommonJS "module.exports" for node compatibility.
+  isNodeMode || !mod || !mod.__esModule ? __defProp(target, "default", { value: mod, enumerable: true }) : target,
+  mod
+));
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var index_exports = {};
+__export(index_exports, {
+  LIBRETTO_WORKFLOW_BRAND: () => import_workflow.LIBRETTO_WORKFLOW_BRAND,
+  LibrettoWorkflow: () => import_workflow.LibrettoWorkflow,
+  Logger: () => import_logger.Logger,
+  SESSION_STATE_VERSION: () => import_state.SESSION_STATE_VERSION,
+  SessionStateFileSchema: () => import_state.SessionStateFileSchema,
+  SessionStatusSchema: () => import_state.SessionStatusSchema,
+  attemptWithRecovery: () => import_recovery.attemptWithRecovery,
+  clearHighlights: () => import_highlight.clearHighlights,
+  createFileLogSink: () => import_sinks.createFileLogSink,
+  createLLMClientFromModel: () => import_ai_sdk_adapter.createLLMClientFromModel,
+  defaultLogger: () => import_logger.defaultLogger,
+  detectSubmissionError: () => import_errors.detectSubmissionError,
+  downloadAndSave: () => import_download.downloadAndSave,
+  downloadViaClick: () => import_download.downloadViaClick,
+  ensureGhostCursor: () => import_ghost_cursor.ensureGhostCursor,
+  ensureHighlightLayer: () => import_highlight.ensureHighlightLayer,
+  executeRecoveryAgent: () => import_agent.executeRecoveryAgent,
+  extractFromPage: () => import_extract.extractFromPage,
+  ghostClick: () => import_ghost_cursor.ghostClick,
+  hideGhostCursor: () => import_ghost_cursor.hideGhostCursor,
+  installInstrumentation: () => import_instrument.installInstrumentation,
+  instrumentContext: () => import_instrument.instrumentContext,
+  instrumentPage: () => import_instrument.instrumentPage,
+  isDebugMode: () => import_config.isDebugMode,
+  isDryRun: () => import_config.isDryRun,
+  jsonlConsoleSink: () => import_sinks.jsonlConsoleSink,
+  launchBrowser: () => import_api.launchBrowser,
+  moveGhostCursor: () => import_ghost_cursor.moveGhostCursor,
+  pageRequest: () => import_network.pageRequest,
+  parseSessionStateContent: () => import_state.parseSessionStateContent,
+  parseSessionStateData: () => import_state.parseSessionStateData,
+  pause: () => import_pause.pause,
+  prettyConsoleSink: () => import_sinks.prettyConsoleSink,
+  serializeSessionState: () => import_state.serializeSessionState,
+  shouldPauseBeforeMutation: () => import_config.shouldPauseBeforeMutation,
+  showHighlight: () => import_highlight.showHighlight,
+  workflow: () => import_workflow.workflow
+});
+module.exports = __toCommonJS(index_exports);
+var import_node_path = require("node:path");
+var import_node_url = require("node:url");
+var import_logger = require("./shared/logger/logger.js");
+var import_sinks = require("./shared/logger/sinks.js");
+var import_ai_sdk_adapter = require("./shared/llm/ai-sdk-adapter.js");
+var import_state = require("./shared/state/index.js");
+var import_agent = require("./runtime/recovery/agent.js");
+var import_recovery = require("./runtime/recovery/recovery.js");
+var import_errors = require("./runtime/recovery/errors.js");
+var import_extract = require("./runtime/extract/extract.js");
+var import_network = require("./runtime/network/network.js");
+var import_download = require("./runtime/download/download.js");
+var import_pause = require("./shared/debug/pause.js");
+var import_config = require("./shared/config/config.js");
+var import_instrument = require("./shared/instrumentation/instrument.js");
+var import_ghost_cursor = require("./shared/visualization/ghost-cursor.js");
+var import_highlight = require("./shared/visualization/highlight.js");
+var import_api = require("./shared/run/api.js");
+var import_workflow = require("./shared/workflow/workflow.js");
+const import_meta = {};
+const isDirectExecution = () => {
+  const entryArg = process.argv[1];
+  if (!entryArg) {
+    return false;
+  }
+  return (0, import_node_url.pathToFileURL)((0, import_node_path.resolve)(entryArg)).href === import_meta.url;
+};
+if (isDirectExecution()) {
+  void import("./cli/index.js").catch((error) => {
+    const message = error instanceof Error ? error.stack ?? error.message : String(error);
+    process.stderr.write(`${message}
+`);
+    process.exitCode = 1;
+  });
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  LIBRETTO_WORKFLOW_BRAND,
+  LibrettoWorkflow,
+  Logger,
+  SESSION_STATE_VERSION,
+  SessionStateFileSchema,
+  SessionStatusSchema,
+  attemptWithRecovery,
+  clearHighlights,
+  createFileLogSink,
+  createLLMClientFromModel,
+  defaultLogger,
+  detectSubmissionError,
+  downloadAndSave,
+  downloadViaClick,
+  ensureGhostCursor,
+  ensureHighlightLayer,
+  executeRecoveryAgent,
+  extractFromPage,
+  ghostClick,
+  hideGhostCursor,
+  installInstrumentation,
+  instrumentContext,
+  instrumentPage,
+  isDebugMode,
+  isDryRun,
+  jsonlConsoleSink,
+  launchBrowser,
+  moveGhostCursor,
+  pageRequest,
+  parseSessionStateContent,
+  parseSessionStateData,
+  pause,
+  prettyConsoleSink,
+  serializeSessionState,
+  shouldPauseBeforeMutation,
+  showHighlight,
+  workflow
+});

package/dist/index.d.cts

@@ -0,0 +1,21 @@
+export { LogOptions, Logger, LoggerApi, LoggerSink, MinimalLogger, defaultLogger } from './shared/logger/logger.cjs';
+export { createFileLogSink, jsonlConsoleSink, prettyConsoleSink } from './shared/logger/sinks.cjs';
+export { LLMClient, Message, MessageContentPart } from './shared/llm/types.cjs';
+export { createLLMClientFromModel } from './shared/llm/ai-sdk-adapter.cjs';
+export { SESSION_STATE_VERSION, SessionState, SessionStateFile, SessionStateFileSchema, SessionStatus, SessionStatusSchema, parseSessionStateContent, parseSessionStateData, serializeSessionState } from './shared/state/session-state.cjs';
+export { executeRecoveryAgent } from './runtime/recovery/agent.cjs';
+export { attemptWithRecovery } from './runtime/recovery/recovery.cjs';
+export { DetectedSubmissionError, KnownSubmissionError, detectSubmissionError } from './runtime/recovery/errors.cjs';
+export { ExtractOptions, extractFromPage } from './runtime/extract/extract.cjs';
+export { PageRequestOptions, RequestConfig, pageRequest } from './runtime/network/network.cjs';
+export { DownloadResult, DownloadViaClickOptions, SaveDownloadOptions, downloadAndSave, downloadViaClick } from './runtime/download/download.cjs';
+export { pause } from './shared/debug/pause.cjs';
+export { isDebugMode, isDryRun, shouldPauseBeforeMutation } from './shared/config/config.cjs';
+export { InstrumentationOptions, InstrumentedPage, installInstrumentation, instrumentContext, instrumentPage } from './shared/instrumentation/instrument.cjs';
+export { GhostCursorOptions, ensureGhostCursor, ghostClick, hideGhostCursor, moveGhostCursor } from './shared/visualization/ghost-cursor.cjs';
+export { HighlightOptions, clearHighlights, ensureHighlightLayer, showHighlight } from './shared/visualization/highlight.cjs';
+export { BrowserSession, LaunchBrowserArgs, launchBrowser } from './shared/run/browser.cjs';
+export { LIBRETTO_WORKFLOW_BRAND, LibrettoWorkflow, LibrettoWorkflowContext, LibrettoWorkflowHandler, LibrettoWorkflowMetadata, workflow } from './shared/workflow/workflow.cjs';
+import 'zod';
+import 'ai';
+import 'playwright';

package/dist/index.d.ts

@@ -0,0 +1,21 @@
+export { LogOptions, Logger, LoggerApi, LoggerSink, MinimalLogger, defaultLogger } from './shared/logger/logger.js';
+export { createFileLogSink, jsonlConsoleSink, prettyConsoleSink } from './shared/logger/sinks.js';
+export { LLMClient, Message, MessageContentPart } from './shared/llm/types.js';
+export { createLLMClientFromModel } from './shared/llm/ai-sdk-adapter.js';
+export { SESSION_STATE_VERSION, SessionState, SessionStateFile, SessionStateFileSchema, SessionStatus, SessionStatusSchema, parseSessionStateContent, parseSessionStateData, serializeSessionState } from './shared/state/session-state.js';
+export { executeRecoveryAgent } from './runtime/recovery/agent.js';
+export { attemptWithRecovery } from './runtime/recovery/recovery.js';
+export { DetectedSubmissionError, KnownSubmissionError, detectSubmissionError } from './runtime/recovery/errors.js';
+export { ExtractOptions, extractFromPage } from './runtime/extract/extract.js';
+export { PageRequestOptions, RequestConfig, pageRequest } from './runtime/network/network.js';
+export { DownloadResult, DownloadViaClickOptions, SaveDownloadOptions, downloadAndSave, downloadViaClick } from './runtime/download/download.js';
+export { pause } from './shared/debug/pause.js';
+export { isDebugMode, isDryRun, shouldPauseBeforeMutation } from './shared/config/config.js';
+export { InstrumentationOptions, InstrumentedPage, installInstrumentation, instrumentContext, instrumentPage } from './shared/instrumentation/instrument.js';
+export { GhostCursorOptions, ensureGhostCursor, ghostClick, hideGhostCursor, moveGhostCursor } from './shared/visualization/ghost-cursor.js';
+export { HighlightOptions, clearHighlights, ensureHighlightLayer, showHighlight } from './shared/visualization/highlight.js';
+export { BrowserSession, LaunchBrowserArgs, launchBrowser } from './shared/run/browser.js';
+export { LIBRETTO_WORKFLOW_BRAND, LibrettoWorkflow, LibrettoWorkflowContext, LibrettoWorkflowHandler, LibrettoWorkflowMetadata, workflow } from './shared/workflow/workflow.js';
+import 'zod';
+import 'ai';
+import 'playwright';

package/dist/index.js

@@ -0,0 +1,114 @@
+import { resolve } from "node:path";
+import { pathToFileURL } from "node:url";
+import { Logger, defaultLogger } from "./shared/logger/logger.js";
+import {
+  createFileLogSink,
+  prettyConsoleSink,
+  jsonlConsoleSink
+} from "./shared/logger/sinks.js";
+import { createLLMClientFromModel } from "./shared/llm/ai-sdk-adapter.js";
+import {
+  SESSION_STATE_VERSION,
+  SessionStatusSchema,
+  SessionStateFileSchema,
+  parseSessionStateData,
+  parseSessionStateContent,
+  serializeSessionState
+} from "./shared/state/index.js";
+import { executeRecoveryAgent } from "./runtime/recovery/agent.js";
+import { attemptWithRecovery } from "./runtime/recovery/recovery.js";
+import {
+  detectSubmissionError
+} from "./runtime/recovery/errors.js";
+import { extractFromPage } from "./runtime/extract/extract.js";
+import {
+  pageRequest
+} from "./runtime/network/network.js";
+import {
+  downloadViaClick,
+  downloadAndSave
+} from "./runtime/download/download.js";
+import { pause } from "./shared/debug/pause.js";
+import {
+  isDebugMode,
+  isDryRun,
+  shouldPauseBeforeMutation
+} from "./shared/config/config.js";
+import {
+  instrumentPage,
+  installInstrumentation,
+  instrumentContext
+} from "./shared/instrumentation/instrument.js";
+import {
+  ensureGhostCursor,
+  moveGhostCursor,
+  ghostClick,
+  hideGhostCursor
+} from "./shared/visualization/ghost-cursor.js";
+import {
+  ensureHighlightLayer,
+  showHighlight,
+  clearHighlights
+} from "./shared/visualization/highlight.js";
+import {
+  launchBrowser
+} from "./shared/run/api.js";
+import {
+  LibrettoWorkflow,
+  LIBRETTO_WORKFLOW_BRAND,
+  workflow
+} from "./shared/workflow/workflow.js";
+const isDirectExecution = () => {
+  const entryArg = process.argv[1];
+  if (!entryArg) {
+    return false;
+  }
+  return pathToFileURL(resolve(entryArg)).href === import.meta.url;
+};
+if (isDirectExecution()) {
+  void import("./cli/index.js").catch((error) => {
+    const message = error instanceof Error ? error.stack ?? error.message : String(error);
+    process.stderr.write(`${message}
+`);
+    process.exitCode = 1;
+  });
+}
+export {
+  LIBRETTO_WORKFLOW_BRAND,
+  LibrettoWorkflow,
+  Logger,
+  SESSION_STATE_VERSION,
+  SessionStateFileSchema,
+  SessionStatusSchema,
+  attemptWithRecovery,
+  clearHighlights,
+  createFileLogSink,
+  createLLMClientFromModel,
+  defaultLogger,
+  detectSubmissionError,
+  downloadAndSave,
+  downloadViaClick,
+  ensureGhostCursor,
+  ensureHighlightLayer,
+  executeRecoveryAgent,
+  extractFromPage,
+  ghostClick,
+  hideGhostCursor,
+  installInstrumentation,
+  instrumentContext,
+  instrumentPage,
+  isDebugMode,
+  isDryRun,
+  jsonlConsoleSink,
+  launchBrowser,
+  moveGhostCursor,
+  pageRequest,
+  parseSessionStateContent,
+  parseSessionStateData,
+  pause,
+  prettyConsoleSink,
+  serializeSessionState,
+  shouldPauseBeforeMutation,
+  showHighlight,
+  workflow
+};

package/dist/runtime/download/download.cjs

@@ -0,0 +1,70 @@
+"use strict";
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var download_exports = {};
+__export(download_exports, {
+  downloadAndSave: () => downloadAndSave,
+  downloadViaClick: () => downloadViaClick
+});
+module.exports = __toCommonJS(download_exports);
+var import_promises = require("node:fs/promises");
+var import_node_path = require("node:path");
+async function downloadViaClick(page, selector, options) {
+  const { logger, timeout = 3e4 } = options ?? {};
+  const startTime = Date.now();
+  const downloadPromise = page.waitForEvent("download", { timeout });
+  await page.locator(selector).click();
+  const download = await downloadPromise;
+  const filename = download.suggestedFilename();
+  const readStream = await download.createReadStream();
+  if (!readStream) {
+    throw new Error(
+      `Download stream unavailable for "${filename}". The browser may have been closed before the download completed.`
+    );
+  }
+  const chunks = [];
+  for await (const chunk of readStream) {
+    chunks.push(Buffer.isBuffer(chunk) ? chunk : Buffer.from(chunk));
+  }
+  const buffer = Buffer.concat(chunks);
+  const duration = Date.now() - startTime;
+  logger?.info("download:click", {
+    selector,
+    filename,
+    size: buffer.length,
+    duration
+  });
+  return { buffer, filename };
+}
+async function downloadAndSave(page, selector, options) {
+  const { savePath, ...downloadOpts } = options ?? {};
+  const { buffer, filename } = await downloadViaClick(page, selector, downloadOpts);
+  const dest = (0, import_node_path.resolve)(savePath ?? filename);
+  await (0, import_promises.writeFile)(dest, buffer);
+  options?.logger?.info("download:saved", {
+    filename,
+    savedTo: dest,
+    size: buffer.length
+  });
+  return { buffer, filename, savedTo: dest };
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  downloadAndSave,
+  downloadViaClick
+});

package/dist/runtime/download/download.d.cts

@@ -0,0 +1,35 @@
+import { Page } from 'playwright';
+import { MinimalLogger } from '../../shared/logger/logger.cjs';
+
+type DownloadResult = {
+    /** The raw file contents. */
+    buffer: Buffer;
+    /** The filename suggested by the server (Content-Disposition header or URL). */
+    filename: string;
+};
+type DownloadViaClickOptions = {
+    logger?: MinimalLogger;
+    /** Timeout in milliseconds for waiting on the download event. Defaults to 30 000. */
+    timeout?: number;
+};
+/**
+ * Triggers a file download by clicking a DOM element and intercepts the
+ * resulting download using Playwright's download event.
+ *
+ * The download promise is registered **before** the click so the event is
+ * never missed.
+ */
+declare function downloadViaClick(page: Page, selector: string, options?: DownloadViaClickOptions): Promise<DownloadResult>;
+type SaveDownloadOptions = DownloadViaClickOptions & {
+    /** Absolute or relative path to save the file to. When omitted the suggested filename is used in the current working directory. */
+    savePath?: string;
+};
+/**
+ * Convenience wrapper around {@link downloadViaClick} that also writes the
+ * downloaded file to disk.
+ */
+declare function downloadAndSave(page: Page, selector: string, options?: SaveDownloadOptions): Promise<DownloadResult & {
+    savedTo: string;
+}>;
+
+export { type DownloadResult, type DownloadViaClickOptions, type SaveDownloadOptions, downloadAndSave, downloadViaClick };

package/dist/runtime/download/download.d.ts

@@ -0,0 +1,35 @@
+import { Page } from 'playwright';
+import { MinimalLogger } from '../../shared/logger/logger.js';
+
+type DownloadResult = {
+    /** The raw file contents. */
+    buffer: Buffer;
+    /** The filename suggested by the server (Content-Disposition header or URL). */
+    filename: string;
+};
+type DownloadViaClickOptions = {
+    logger?: MinimalLogger;
+    /** Timeout in milliseconds for waiting on the download event. Defaults to 30 000. */
+    timeout?: number;
+};
+/**
+ * Triggers a file download by clicking a DOM element and intercepts the
+ * resulting download using Playwright's download event.
+ *
+ * The download promise is registered **before** the click so the event is
+ * never missed.
+ */
+declare function downloadViaClick(page: Page, selector: string, options?: DownloadViaClickOptions): Promise<DownloadResult>;
+type SaveDownloadOptions = DownloadViaClickOptions & {
+    /** Absolute or relative path to save the file to. When omitted the suggested filename is used in the current working directory. */
+    savePath?: string;
+};
+/**
+ * Convenience wrapper around {@link downloadViaClick} that also writes the
+ * downloaded file to disk.
+ */
+declare function downloadAndSave(page: Page, selector: string, options?: SaveDownloadOptions): Promise<DownloadResult & {
+    savedTo: string;
+}>;
+
+export { type DownloadResult, type DownloadViaClickOptions, type SaveDownloadOptions, downloadAndSave, downloadViaClick };

package/dist/runtime/download/download.js

@@ -0,0 +1,45 @@
+import { writeFile } from "node:fs/promises";
+import { resolve } from "node:path";
+async function downloadViaClick(page, selector, options) {
+  const { logger, timeout = 3e4 } = options ?? {};
+  const startTime = Date.now();
+  const downloadPromise = page.waitForEvent("download", { timeout });
+  await page.locator(selector).click();
+  const download = await downloadPromise;
+  const filename = download.suggestedFilename();
+  const readStream = await download.createReadStream();
+  if (!readStream) {
+    throw new Error(
+      `Download stream unavailable for "${filename}". The browser may have been closed before the download completed.`
+    );
+  }
+  const chunks = [];
+  for await (const chunk of readStream) {
+    chunks.push(Buffer.isBuffer(chunk) ? chunk : Buffer.from(chunk));
+  }
+  const buffer = Buffer.concat(chunks);
+  const duration = Date.now() - startTime;
+  logger?.info("download:click", {
+    selector,
+    filename,
+    size: buffer.length,
+    duration
+  });
+  return { buffer, filename };
+}
+async function downloadAndSave(page, selector, options) {
+  const { savePath, ...downloadOpts } = options ?? {};
+  const { buffer, filename } = await downloadViaClick(page, selector, downloadOpts);
+  const dest = resolve(savePath ?? filename);
+  await writeFile(dest, buffer);
+  options?.logger?.info("download:saved", {
+    filename,
+    savedTo: dest,
+    size: buffer.length
+  });
+  return { buffer, filename, savedTo: dest };
+}
+export {
+  downloadAndSave,
+  downloadViaClick
+};

package/dist/runtime/download/index.cjs

@@ -0,0 +1,30 @@
+"use strict";
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var download_exports = {};
+__export(download_exports, {
+  downloadAndSave: () => import_download.downloadAndSave,
+  downloadViaClick: () => import_download.downloadViaClick
+});
+module.exports = __toCommonJS(download_exports);
+var import_download = require("./download.js");
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  downloadAndSave,
+  downloadViaClick
+});

package/dist/runtime/download/index.d.cts

@@ -0,0 +1,3 @@
+export { DownloadResult, DownloadViaClickOptions, SaveDownloadOptions, downloadAndSave, downloadViaClick } from './download.cjs';
+import 'playwright';
+import '../../shared/logger/logger.cjs';

package/dist/runtime/download/index.d.ts

@@ -0,0 +1,3 @@
+export { DownloadResult, DownloadViaClickOptions, SaveDownloadOptions, downloadAndSave, downloadViaClick } from './download.js';
+import 'playwright';
+import '../../shared/logger/logger.js';

package/dist/runtime/download/index.js

@@ -0,0 +1,8 @@
+import {
+  downloadViaClick,
+  downloadAndSave
+} from "./download.js";
+export {
+  downloadAndSave,
+  downloadViaClick
+};

package/dist/runtime/extract/extract.cjs

@@ -0,0 +1,88 @@
+"use strict";
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var extract_exports = {};
+__export(extract_exports, {
+  extractFromPage: () => extractFromPage
+});
+module.exports = __toCommonJS(extract_exports);
+var import_logger = require("../../shared/logger/logger.js");
+async function extractFromPage(options) {
+  const { page, instruction, schema, selector, logger = import_logger.defaultLogger, llmClient } = options;
+  let screenshot;
+  let domContent;
+  if (selector) {
+    const element = page.locator(selector);
+    await element.waitFor({ state: "visible", timeout: 1e4 });
+    const screenshotBuffer = await element.screenshot();
+    screenshot = screenshotBuffer.toString("base64");
+    try {
+      domContent = await element.innerHTML();
+      if (domContent.length > 3e4) {
+        domContent = domContent.slice(0, 3e4) + "\n... [truncated]";
+      }
+    } catch {
+      domContent = void 0;
+    }
+  } else {
+    const cdpClient = await page.context().newCDPSession(page);
+    await cdpClient.send("Page.enable");
+    const { data } = await cdpClient.send("Page.captureScreenshot", {
+      format: "png"
+    });
+    screenshot = data;
+    try {
+      const htmlContent = await page.content();
+      domContent = htmlContent.length > 5e4 ? htmlContent.slice(0, 5e4) + "\n... [truncated]" : htmlContent;
+    } catch {
+      domContent = void 0;
+    }
+  }
+  const prompt = `You are analyzing a screenshot${selector ? " of a specific element" : ""} from a web page to extract structured data.
+
+Instruction: ${instruction}
+
+${domContent ? `Here is the HTML content for additional context:
+<html>
+${domContent}
+</html>` : ""}
+
+Extract the requested information from the screenshot and return it in the specified format. Be precise and only extract what is visible.`;
+  const result = await llmClient.generateObjectFromMessages({
+    schema,
+    messages: [
+      {
+        role: "user",
+        content: [
+          { type: "text", text: prompt },
+          { type: "image", image: `data:image/png;base64,${screenshot}` }
+        ]
+      }
+    ],
+    temperature: 0
+  });
+  logger.info("extractFromPage completed", {
+    selector,
+    instruction: instruction.slice(0, 100)
+  });
+  return result;
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  extractFromPage
+});

package/dist/runtime/extract/extract.d.cts

@@ -0,0 +1,23 @@
+import { Page } from 'playwright';
+import z from 'zod';
+import { MinimalLogger } from '../../shared/logger/logger.cjs';
+import { LLMClient } from '../../shared/llm/types.cjs';
+
+type ExtractOptions<T extends z.ZodType> = {
+    page: Page;
+    instruction: string;
+    schema: T;
+    llmClient: LLMClient;
+    logger?: MinimalLogger;
+    /** Optional CSS selector to scope extraction to a specific element. */
+    selector?: string;
+};
+/**
+ * Generic AI-powered data extraction from page elements.
+ * Takes a screenshot (full-page via CDP or scoped to an element),
+ * captures DOM content, and uses an LLM to extract structured data
+ * matching the provided Zod schema.
+ */
+declare function extractFromPage<T extends z.ZodType>(options: ExtractOptions<T>): Promise<z.infer<T>>;
+
+export { type ExtractOptions, extractFromPage };