mcp-scraper 0.2.15 → 0.2.16

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (33) hide show
  1. package/README.md +4 -4
  2. package/dist/bin/api-server.cjs +11 -10
  3. package/dist/bin/api-server.cjs.map +1 -1
  4. package/dist/bin/api-server.js +1 -1
  5. package/dist/bin/browser-agent-stdio-server.cjs +1 -1
  6. package/dist/bin/browser-agent-stdio-server.cjs.map +1 -1
  7. package/dist/bin/browser-agent-stdio-server.js +2 -2
  8. package/dist/bin/mcp-scraper-cli.cjs +1 -1
  9. package/dist/bin/mcp-scraper-cli.cjs.map +1 -1
  10. package/dist/bin/mcp-scraper-cli.js +1 -1
  11. package/dist/bin/mcp-scraper-combined-stdio-server.cjs +10 -9
  12. package/dist/bin/mcp-scraper-combined-stdio-server.cjs.map +1 -1
  13. package/dist/bin/mcp-scraper-combined-stdio-server.js +3 -3
  14. package/dist/bin/mcp-scraper-install.cjs +3 -3
  15. package/dist/bin/mcp-scraper-install.cjs.map +1 -1
  16. package/dist/bin/mcp-scraper-install.js +3 -3
  17. package/dist/bin/mcp-scraper-install.js.map +1 -1
  18. package/dist/bin/mcp-stdio-server.cjs +10 -9
  19. package/dist/bin/mcp-stdio-server.cjs.map +1 -1
  20. package/dist/bin/mcp-stdio-server.js +2 -2
  21. package/dist/{chunk-TIPUIEJN.js → chunk-ATJAINML.js} +2 -2
  22. package/dist/{chunk-Q4DFONIK.js → chunk-FMC3V54I.js} +11 -10
  23. package/dist/chunk-FMC3V54I.js.map +1 -0
  24. package/dist/chunk-KE4JZDLV.js +7 -0
  25. package/dist/chunk-KE4JZDLV.js.map +1 -0
  26. package/dist/{server-VD2TD3AD.js → server-KSEQLZNP.js} +4 -4
  27. package/dist/server-KSEQLZNP.js.map +1 -0
  28. package/package.json +1 -1
  29. package/dist/chunk-2CQXHSWC.js +0 -7
  30. package/dist/chunk-2CQXHSWC.js.map +0 -1
  31. package/dist/chunk-Q4DFONIK.js.map +0 -1
  32. package/dist/server-VD2TD3AD.js.map +0 -1
  33. /package/dist/{chunk-TIPUIEJN.js.map → chunk-ATJAINML.js.map} +0 -0
@@ -2,9 +2,9 @@
2
2
  import {
3
3
  HttpMcpToolExecutor,
4
4
  buildPaaExtractorMcpServer
5
- } from "../chunk-Q4DFONIK.js";
5
+ } from "../chunk-FMC3V54I.js";
6
6
  import "../chunk-M2S27J6Z.js";
7
- import "../chunk-2CQXHSWC.js";
7
+ import "../chunk-KE4JZDLV.js";
8
8
 
9
9
  // bin/mcp-stdio-server.ts
10
10
  import { readFileSync } from "fs";
@@ -1,6 +1,6 @@
1
1
  import {
2
2
  PACKAGE_VERSION
3
- } from "./chunk-2CQXHSWC.js";
3
+ } from "./chunk-KE4JZDLV.js";
4
4
 
5
5
  // src/mcp/browser-agent-mcp-server.ts
6
6
  import { McpServer } from "@modelcontextprotocol/sdk/server/mcp.js";
@@ -790,4 +790,4 @@ export {
790
790
  buildBrowserAgentMcpServer,
791
791
  registerBrowserAgentMcpTools
792
792
  };
793
- //# sourceMappingURL=chunk-TIPUIEJN.js.map
793
+ //# sourceMappingURL=chunk-ATJAINML.js.map
@@ -3,7 +3,7 @@ import {
3
3
  } from "./chunk-M2S27J6Z.js";
4
4
  import {
5
5
  PACKAGE_VERSION
6
- } from "./chunk-2CQXHSWC.js";
6
+ } from "./chunk-KE4JZDLV.js";
7
7
 
8
8
  // src/harvest-timeout.ts
9
9
  var VERCEL_FUNCTION_MAX_MS = 3e5;
@@ -570,7 +570,8 @@ ${adBlocks}`,
570
570
  `
571
571
  ---
572
572
  \u{1F4A1} **Tips**
573
- - Transcribe video ads: use \`facebook_ad_transcribe\` with the \`videoUrl\` above
573
+ - Transcribe video ads: use \`facebook_ad_transcribe\` with the direct \`videoUrl\` above
574
+ - Transcribe organic Facebook reels/posts: use \`facebook_video_transcribe\` with the public Facebook URL
574
575
  - Find other advertisers: use \`facebook_ad_search\``
575
576
  ].filter(Boolean).join("\n");
576
577
  return {
@@ -1016,7 +1017,7 @@ ${text}`,
1016
1017
  ${chunkRows}` : "",
1017
1018
  `
1018
1019
  ---
1019
- \u{1F4A1} Get more ads from this advertiser: use \`facebook_page_intel\``
1020
+ \u{1F4A1} Get more ads from this advertiser: use \`facebook_page_intel\`. For public Facebook reel/post URLs, use \`facebook_video_transcribe\`.`
1020
1021
  ].filter(Boolean).join("\n");
1021
1022
  return oneBlock(full);
1022
1023
  }
@@ -1052,7 +1053,7 @@ ${text}`,
1052
1053
  ${chunkRows}` : "",
1053
1054
  `
1054
1055
  ---
1055
- \u{1F4A1} To download the extracted MP4, call the Facebook media endpoint with the extracted MP4 URL.`
1056
+ \u{1F4A1} Use \`videoUrl\` as the extracted MP4 for download or follow-up processing. Use \`facebook_ad_transcribe\` only for Ad Library videoUrl values.`
1056
1057
  ].filter(Boolean).join("\n");
1057
1058
  return {
1058
1059
  ...oneBlock(full),
@@ -1129,10 +1130,10 @@ var FacebookAdSearchInputSchema = {
1129
1130
  maxResults: z.number().int().min(1).max(20).default(10)
1130
1131
  };
1131
1132
  var FacebookAdTranscribeInputSchema = {
1132
- videoUrl: z.string().url().describe("Facebook CDN video URL from a facebook_page_intel result")
1133
+ videoUrl: z.string().url().describe("Direct Facebook CDN video URL from a facebook_page_intel ad result. Do not pass a public Facebook reel/post/share URL here; use facebook_video_transcribe for organic Facebook URLs.")
1133
1134
  };
1134
1135
  var FacebookVideoTranscribeInputSchema = {
1135
- url: z.string().url().describe("Organic Facebook reel, video, watch, or share URL. The tool renders the page, extracts the best public Facebook CDN MP4 URL, then transcribes it."),
1136
+ url: z.string().url().describe("Organic Facebook reel, video, watch, post, or share URL from facebook.com, m.facebook.com, or fb.watch. The tool renders the page, extracts the best matching public Facebook CDN MP4 URL, then transcribes it. Use this when the user pastes a normal Facebook video page URL and asks for the transcript or downloadable MP4."),
1136
1137
  quality: z.enum(["best", "hd", "sd"]).default("best").describe("Preferred progressive MP4 quality. Use best by default; hd prefers the highest HD progressive URL; sd forces the SD URL.")
1137
1138
  };
1138
1139
  var MapsPlaceIntelInputSchema = {
@@ -1999,7 +2000,7 @@ function registerPaaExtractorMcpTools(server, executor, options = {}) {
1999
2000
  }, async (input) => formatYoutubeTranscribe(await executor.youtubeTranscribe(input), input));
2000
2001
  server.registerTool("facebook_page_intel", {
2001
2002
  title: "Facebook Advertiser Ad Intel",
2002
- description: withReportNote("Harvest ads from a Facebook advertiser. Returns ad copy, headlines, CTAs, creative type, status, landing URLs, and video URLs ready for transcription. Accepts pageId, libraryId, or a brand/advertiser name as query. Use after facebook_ad_search when possible."),
2003
+ description: withReportNote("Harvest ads from a Facebook advertiser. Returns ad copy, headlines, CTAs, creative type, status, landing URLs, and direct ad video URLs ready for facebook_ad_transcribe. Accepts pageId, libraryId, or a brand/advertiser name as query. Use after facebook_ad_search when possible. For normal public Facebook reels/posts/watch/share URLs, use facebook_video_transcribe instead."),
2003
2004
  inputSchema: FacebookPageIntelInputSchema,
2004
2005
  outputSchema: FacebookPageIntelOutputSchema,
2005
2006
  annotations: liveWebToolAnnotations("Facebook Advertiser Ad Intel")
@@ -2013,13 +2014,13 @@ function registerPaaExtractorMcpTools(server, executor, options = {}) {
2013
2014
  }, async (input) => formatFacebookAdSearch(await executor.facebookAdSearch(input), input));
2014
2015
  server.registerTool("facebook_ad_transcribe", {
2015
2016
  title: "Facebook Ad Transcription",
2016
- description: "Transcribe audio from a Facebook ad video. Returns full transcript and timestamped chunks. Use the videoUrl value from facebook_page_intel results.",
2017
+ description: "Transcribe audio from a Facebook ad video CDN URL. Returns full transcript and timestamped chunks. Use only with the direct videoUrl value from facebook_page_intel results, not public Facebook post/reel/share URLs.",
2017
2018
  inputSchema: FacebookAdTranscribeInputSchema,
2018
2019
  annotations: liveWebToolAnnotations("Facebook Ad Transcription")
2019
2020
  }, async (input) => formatFacebookAdTranscribe(await executor.facebookAdTranscribe(input), input));
2020
2021
  server.registerTool("facebook_video_transcribe", {
2021
2022
  title: "Facebook Organic Video Transcription",
2022
- description: withReportNote("Transcribe audio from an organic Facebook reel, video, watch, or share URL. Renders the Facebook page, extracts the best public Facebook CDN MP4 URL from the page state, then returns the full transcript, timestamped chunks, and extracted MP4 URL."),
2023
+ description: withReportNote("Transcribe audio from an organic Facebook reel, video, watch, post, or share URL, including fb.watch links. Use this when the user pastes a normal Facebook video page URL and wants the transcript or downloadable MP4. Renders the Facebook page in a browser, selects the best matching public Facebook CDN MP4 URL from page state, then returns sourceUrl, resolved pageUrl, videoId, ownerName, selectedQuality, bitrate, videoDurationSec, extracted MP4 URL, full transcript, and timestamped chunks."),
2023
2024
  inputSchema: FacebookVideoTranscribeInputSchema,
2024
2025
  outputSchema: FacebookVideoTranscribeOutputSchema,
2025
2026
  annotations: liveWebToolAnnotations("Facebook Organic Video Transcription")
@@ -2190,4 +2191,4 @@ export {
2190
2191
  registerPaaExtractorMcpTools,
2191
2192
  HttpMcpToolExecutor
2192
2193
  };
2193
- //# sourceMappingURL=chunk-Q4DFONIK.js.map
2194
+ //# sourceMappingURL=chunk-FMC3V54I.js.map