@speakai/mcp-server 1.11.0 → 1.12.1

package/README.md

@@ -402,9 +402,9 @@ SPEAK_API_KEY=your-key npx @speakai/mcp-server
 | Tool | Description |
 |---|---|
 | `get_signed_upload_url` | Get a pre-signed S3 URL for direct file upload |
-| `upload_media` | Upload media from a public URL for transcription |
+| `upload_media` | Upload media from a URL — a direct/public file URL or a social/video link (YouTube, Instagram, TikTok, X, Facebook, Reddit, SoundCloud, …) resolved automatically. |
 | `upload_local_file` | Upload a local file directly from disk |
-| `upload_and_analyze` | Upload media and return its `media_id` immediately. Poll `get_media_status` until `processed`, then call `get_media_insights` for AI summaries. |
+| `upload_and_analyze` | Upload media from a URL (direct file or social/video link, resolved automatically) and return its `media_id` immediately. Poll `get_media_status` until `processed`, then call `get_media_insights` for AI summaries. |
 | `list_media` | List and search media files with filters, pagination, and optional inline data (transcripts, speakers, keywords) via `include` param |
 | `get_media_insights` | Get AI insights — topics, sentiment, summaries, action items |
 | `get_transcript` | Get full transcript with speaker labels and timestamps |

package/dist/index.js

@@ -1208,10 +1208,10 @@ function register(server, client) {
   registerSpeakTool(
     server,
     "upload_media",
-    "Upload media from a publicly accessible URL. Processing is asynchronous \u2014 after uploading, use get_media_status to poll until state is 'processed' (typically 1-3 minutes for audio under 60 min), then use get_transcript and get_media_insights to retrieve results. For a single call that handles everything, use upload_and_analyze instead. For local files, use upload_local_file.",
+    "Upload media from a URL \u2014 a direct/public file URL, a pre-signed S3 URL, or a shareable social/video link (YouTube, Instagram, TikTok, X, Facebook, Reddit, SoundCloud, and similar) which Speak resolves to the underlying media automatically. Processing is asynchronous \u2014 after uploading, use get_media_status to poll until state is 'processed' (typically 1-3 minutes for audio under 60 min), then use get_transcript and get_media_insights to retrieve results. For a single call that handles everything, use upload_and_analyze instead. For local files, use upload_local_file. (Vimeo links are not yet supported.)",
     {
       name: import_zod2.z.string().min(1).describe("Display name for the media file"),
-      url: import_zod2.z.string().describe("Publicly accessible URL of the media file (or pre-signed S3 URL)"),
+      url: import_zod2.z.string().describe("Direct/public media file URL, pre-signed S3 URL, or a shareable social/video page link (e.g. an Instagram reel or TikTok URL) \u2014 page links are resolved to the underlying media server-side."),
       mediaType: import_zod2.z.enum([MediaType.AUDIO, MediaType.VIDEO]).describe('Type of media: "audio" or "video"'),
       description: import_zod2.z.string().optional().describe("Description of the media file"),
       sourceLanguage: import_zod2.z.string().optional().describe('BCP-47 language code for transcription, e.g. "en-US" or "he-IL"'),
@@ -4214,9 +4214,9 @@ function register14(server, client) {
   registerSpeakTool(
     server,
     "upload_and_analyze",
-    "Upload media and return media_id immediately. After this returns, poll get_media_status until state is 'processed' (typically 1-3 min for under 60min audio), then call get_media_insights for AI summaries. This async pattern is required for remote MCP transports \u2014 long blocking calls die at proxy idle timeouts.",
+    "Upload and transcribe media from a URL \u2014 a direct/public file URL, OR a shareable social/video link (YouTube, Instagram, TikTok, X, Facebook, Reddit, SoundCloud, and similar), which Speak resolves to the underlying media automatically. Returns media_id immediately; after this returns, poll get_media_status until state is 'processed' (typically 1-3 min for under 60min audio), then call get_media_insights for AI summaries. This async pattern is required for remote MCP transports \u2014 long blocking calls die at proxy idle timeouts. (Vimeo links are not yet supported.)",
     {
-      url: import_zod15.z.string().describe("Publicly accessible URL of the media file"),
+      url: import_zod15.z.string().describe("Direct/public media file URL, or a shareable social/video page link (e.g. an Instagram reel, TikTok, YouTube, or X post URL) \u2014 page links are resolved to the underlying media server-side. Pass the URL the user gave you as-is."),
       name: import_zod15.z.string().optional().describe("Display name for the media (defaults to filename from URL)"),
       mediaType: import_zod15.z.enum([MediaType.AUDIO, MediaType.VIDEO]).optional().describe("Media type (default: audio)"),
       sourceLanguage: import_zod15.z.string().optional().describe("BCP-47 language code (e.g., 'en-US', 'he-IL')"),
@@ -4540,7 +4540,7 @@ function registerPrompts(server) {
     "analyze-meeting",
     "Upload a meeting recording and get a full analysis \u2014 transcript, insights, action items, and key takeaways.",
     {
-      url: import_zod16.z.string().describe("URL of the meeting recording"),
+      url: import_zod16.z.string().describe("URL of the meeting recording \u2014 a direct file link or a shareable social/video link (resolved automatically)"),
       name: import_zod16.z.string().optional().describe("Meeting name (optional)")
     },
     async ({ url, name }) => ({

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@speakai/mcp-server",
-  "version": "1.11.0",
+  "version": "1.12.1",
   "mcpName": "io.github.speakai/mcp-server",
   "description": "Official Speak AI MCP Server — capture meetings, search thousands of recordings, run async voice and video surveys, create clips, and automate workflows from your AI assistant.",
   "homepage": "https://mcp.speakai.co",