mcp-video-analyzer 0.2.3 → 0.2.4

package/README.md

@@ -1,5 +1,11 @@
 # mcp-video-analyzer
 
+<a href="https://glama.ai/mcp/servers/guimatheus92/mcp-video-analyzer">
+  <img width="380" height="200" src="https://glama.ai/mcp/servers/guimatheus92/mcp-video-analyzer/badge" alt="mcp-video-analyzer MCP server" />
+</a>
+
+Featured in [awesome-mcp-servers](https://github.com/punkpeye/awesome-mcp-servers#-multimedia-process).
+
 MCP server for video analysis — extracts transcripts, key frames, and metadata from video URLs. Supports Loom, direct video files (.mp4, .webm), and more.
 
 No existing video MCP combines **transcripts + visual frames + metadata** in one tool. This one does.

package/dist/processors/frame-ocr.d.ts

@@ -10,4 +10,4 @@ export interface IOcrResult {
  *
  * Only includes results with meaningful text (confidence > 50%, text length > 3).
  */
-export declare function extractTextFromFrames(frames: IFrameResult[]): Promise<IOcrResult[]>;
+export declare function extractTextFromFrames(frames: IFrameResult[], language?: string): Promise<IOcrResult[]>;

package/dist/processors/frame-ocr.js

@@ -4,11 +4,11 @@
  *
  * Only includes results with meaningful text (confidence > 50%, text length > 3).
  */
-export async function extractTextFromFrames(frames) {
+export async function extractTextFromFrames(frames, language = 'eng+por') {
     const Tesseract = await loadTesseract();
     if (!Tesseract)
         return [];
-    const worker = await Tesseract.createWorker('eng');
+    const worker = await Tesseract.createWorker(language);
     try {
         const results = [];
         for (const frame of frames) {

package/dist/server.js

@@ -12,7 +12,7 @@ import { registerGetTranscript } from './tools/get-transcript.js';
 export function createServer() {
     const server = new FastMCP({
         name: 'mcp-video-analyzer',
-        version: '0.2.3',
+        version: '0.2.4',
         instructions: `Video analysis MCP server. Extracts transcripts, key frames, metadata, comments, OCR text, and annotated timelines from video URLs.
 
 AUTOMATIC BEHAVIOR — Do NOT wait for the user to ask:

package/dist/tools/analyze-moment.js

@@ -18,6 +18,10 @@ const AnalyzeMomentSchema = z.object({
         .default(10)
         .optional()
         .describe('Number of frames to extract in the range (default: 10)'),
+    ocrLanguage: z
+        .string()
+        .optional()
+        .describe('Tesseract OCR language codes (default: "eng+por"). Use "+" to combine: "eng+spa", "eng+fra+deu".'),
 });
 export function registerAnalyzeMoment(server) {
     server.addTool({
@@ -47,6 +51,7 @@ Requires video download capability for frame extraction.`,
         execute: async (args, { reportProgress }) => {
             const { url, from, to } = args;
             const count = args.count ?? 10;
+            const ocrLanguage = args.ocrLanguage ?? 'eng+por';
             // Validate timestamps
             const fromSeconds = parseTimestamp(from);
             const toSeconds = parseTimestamp(to);
@@ -103,7 +108,7 @@ Requires video download capability for frame extraction.`,
             }
             await reportProgress({ progress: 75, total: 100 });
             // OCR
-            const ocrResults = await extractTextFromFrames(frames).catch((e) => {
+            const ocrResults = await extractTextFromFrames(frames, ocrLanguage).catch((e) => {
                 warnings.push(`OCR failed: ${e instanceof Error ? e.message : String(e)}`);
                 return [];
             });

package/dist/tools/analyze-video.js

@@ -62,6 +62,10 @@ const AnalyzeOptionsSchema = z
         .default(false)
         .optional()
         .describe('Bypass cache and re-analyze the video'),
+    ocrLanguage: z
+        .string()
+        .optional()
+        .describe('Tesseract OCR language codes (default: "eng+por"). Use "+" to combine: "eng+spa", "eng+fra+deu". See Tesseract docs for codes.'),
 })
     .optional();
 const AnalyzeVideoSchema = z.object({
@@ -104,6 +108,7 @@ Use options.forceRefresh to bypass the cache.`,
             const forceRefresh = options?.forceRefresh ?? false;
             const fields = options?.fields;
             const threshold = options?.threshold ?? 0.1;
+            const ocrLanguage = options?.ocrLanguage ?? 'eng+por';
             // Resolve detail config
             const config = getDetailConfig(detail);
             const maxFrames = options?.maxFrames ?? config.maxFrames;
@@ -267,7 +272,7 @@ Use options.forceRefresh to bypass the cache.`,
                         await reportProgress({ progress: 85, total: 100 });
                         // OCR: extract text visible on screen
                         if (config.includeOcr) {
-                            result.ocrResults = await extractTextFromFrames(result.frames).catch((e) => {
+                            result.ocrResults = await extractTextFromFrames(result.frames, ocrLanguage).catch((e) => {
                                 warnings.push(`OCR failed: ${e instanceof Error ? e.message : String(e)}`);
                                 return [];
                             });

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "mcp-video-analyzer",
-  "version": "0.2.3",
+  "version": "0.2.4",
   "description": "MCP server for video analysis — extracts transcripts, key frames, OCR text, and metadata from video URLs. Supports Loom and direct video files.",
   "author": "guimatheus92",
   "license": "MIT",