@juspay/neurolink 9.5.3 → 9.7.0

package/dist/processors/config/mimeTypes.js

@@ -313,6 +313,31 @@ export const EXTENSION_MIME_MAP = {
     ".lua": "text/x-lua",
     ".perl": "text/x-perl",
     ".pl": "text/x-perl",
+    // Video
+    ".mp4": VIDEO_MIME_TYPES.MP4,
+    ".m4v": VIDEO_MIME_TYPES.MP4,
+    ".mkv": VIDEO_MIME_TYPES.MKV,
+    ".mov": VIDEO_MIME_TYPES.MOV,
+    ".avi": VIDEO_MIME_TYPES.AVI,
+    ".webm": VIDEO_MIME_TYPES.WEBM,
+    ".wmv": VIDEO_MIME_TYPES.WMV,
+    ".flv": VIDEO_MIME_TYPES.FLV,
+    // Audio
+    ".mp3": AUDIO_MIME_TYPES.MP3,
+    ".wav": AUDIO_MIME_TYPES.WAV,
+    ".ogg": AUDIO_MIME_TYPES.OGG,
+    ".flac": AUDIO_MIME_TYPES.FLAC,
+    ".m4a": AUDIO_MIME_TYPES.M4A,
+    ".aac": AUDIO_MIME_TYPES.AAC,
+    ".wma": AUDIO_MIME_TYPES.WMA,
+    // Archives
+    ".zip": ARCHIVE_MIME_TYPES.ZIP,
+    ".tar": ARCHIVE_MIME_TYPES.TAR,
+    ".gz": ARCHIVE_MIME_TYPES.GZIP,
+    ".tgz": ARCHIVE_MIME_TYPES.GZIP,
+    ".rar": ARCHIVE_MIME_TYPES.RAR,
+    ".7z": ARCHIVE_MIME_TYPES.SEVEN_ZIP,
+    ".jar": "application/java-archive",
     // Config files
     ".env": TEXT_MIME_TYPES.PLAIN,
     ".ini": TEXT_MIME_TYPES.PLAIN,

package/dist/processors/config/sizeLimits.d.ts

@@ -10,28 +10,34 @@
 export declare const SIZE_LIMITS_MB: {
     /** Maximum image file size (10MB) */
     readonly IMAGE_MAX_MB: 10;
-    /** Maximum PDF file size (20MB) */
-    readonly PDF_MAX_MB: 20;
-    /** Maximum document file size (10MB) */
-    readonly DOCUMENT_MAX_MB: 10;
-    /** Maximum Word document size (10MB) */
-    readonly WORD_MAX_MB: 10;
-    /** Maximum text file size (5MB) */
-    readonly TEXT_MAX_MB: 5;
-    /** Maximum CSV file size (10MB) */
-    readonly CSV_MAX_MB: 10;
-    /** Maximum Excel file size (10MB) */
-    readonly EXCEL_MAX_MB: 10;
-    /** Maximum source code file size (5MB) */
-    readonly SOURCE_CODE_MAX_MB: 5;
-    /** Maximum JSON file size (5MB) */
-    readonly JSON_MAX_MB: 5;
-    /** Maximum YAML file size (5MB) */
-    readonly YAML_MAX_MB: 5;
-    /** Maximum XML file size (5MB) */
-    readonly XML_MAX_MB: 5;
-    /** Maximum general file size (50MB) */
-    readonly GENERAL_MAX_MB: 50;
+    /** Maximum PDF file size (100MB — enterprise reports, long research papers) */
+    readonly PDF_MAX_MB: 100;
+    /** Maximum document file size (100MB — enterprise docs with embedded images) */
+    readonly DOCUMENT_MAX_MB: 100;
+    /** Maximum Word document size (100MB — PRDs, specs with embedded media) */
+    readonly WORD_MAX_MB: 100;
+    /** Maximum text file size (50MB — large log files, build output) */
+    readonly TEXT_MAX_MB: 50;
+    /** Maximum CSV file size (50MB — large data exports) */
+    readonly CSV_MAX_MB: 50;
+    /** Maximum Excel file size (100MB — enterprise spreadsheets, data exports) */
+    readonly EXCEL_MAX_MB: 100;
+    /** Maximum source code file size (20MB — large codebases) */
+    readonly SOURCE_CODE_MAX_MB: 20;
+    /** Maximum JSON file size (20MB — large API response dumps) */
+    readonly JSON_MAX_MB: 20;
+    /** Maximum YAML file size (20MB — large config files) */
+    readonly YAML_MAX_MB: 20;
+    /** Maximum XML file size (20MB — large config/data files) */
+    readonly XML_MAX_MB: 20;
+    /** Maximum video file size (500MB — long meeting recordings, screen captures) */
+    readonly VIDEO_MAX_MB: 500;
+    /** Maximum audio file size (500MB — long meeting recordings, call recordings) */
+    readonly AUDIO_MAX_MB: 500;
+    /** Maximum archive file size (200MB — large project archives) */
+    readonly ARCHIVE_MAX_MB: 200;
+    /** Maximum general file size (200MB — matches Curator's memory-safety cap) */
+    readonly GENERAL_MAX_MB: 200;
 };
 /**
  * Maximum file sizes in bytes for different file types
@@ -39,27 +45,33 @@ export declare const SIZE_LIMITS_MB: {
 export declare const SIZE_LIMITS_BYTES: {
     /** Maximum image file size (10MB) */
     readonly IMAGE_MAX: number;
-    /** Maximum PDF file size (20MB) */
+    /** Maximum PDF file size (100MB) */
     readonly PDF_MAX: number;
-    /** Maximum document file size (10MB) */
+    /** Maximum document file size (100MB) */
     readonly DOCUMENT_MAX: number;
-    /** Maximum Word document size (10MB) */
+    /** Maximum Word document size (100MB) */
     readonly WORD_MAX: number;
-    /** Maximum text file size (5MB) */
+    /** Maximum text file size (50MB) */
     readonly TEXT_MAX: number;
-    /** Maximum CSV file size (10MB) */
+    /** Maximum CSV file size (50MB) */
     readonly CSV_MAX: number;
-    /** Maximum Excel file size (10MB) */
+    /** Maximum Excel file size (100MB) */
     readonly EXCEL_MAX: number;
-    /** Maximum source code file size (5MB) */
+    /** Maximum source code file size (20MB) */
     readonly SOURCE_CODE_MAX: number;
-    /** Maximum JSON file size (5MB) */
+    /** Maximum JSON file size (20MB) */
     readonly JSON_MAX: number;
-    /** Maximum YAML file size (5MB) */
+    /** Maximum YAML file size (20MB) */
     readonly YAML_MAX: number;
-    /** Maximum XML file size (5MB) */
+    /** Maximum XML file size (20MB) */
     readonly XML_MAX: number;
-    /** Maximum general file size (50MB) */
+    /** Maximum video file size (500MB) */
+    readonly VIDEO_MAX: number;
+    /** Maximum audio file size (500MB) */
+    readonly AUDIO_MAX: number;
+    /** Maximum archive file size (200MB) */
+    readonly ARCHIVE_MAX: number;
+    /** Maximum general file size (200MB) */
     readonly GENERAL_MAX: number;
 };
 /**
@@ -114,13 +126,13 @@ export declare const YAML_LIMITS: {
  */
 export declare const SIZE_LIMITS: {
     readonly IMAGE_MAX_MB: 10;
-    readonly PDF_MAX_MB: 20;
-    readonly DOCUMENT_MAX_MB: 10;
-    readonly WORD_MAX_MB: 10;
-    readonly TEXT_MAX_MB: 5;
-    readonly CSV_MAX_MB: 10;
-    readonly EXCEL_MAX_MB: 10;
-    readonly SOURCE_CODE_MAX_MB: 5;
+    readonly PDF_MAX_MB: 100;
+    readonly DOCUMENT_MAX_MB: 100;
+    readonly WORD_MAX_MB: 100;
+    readonly TEXT_MAX_MB: 50;
+    readonly CSV_MAX_MB: 50;
+    readonly EXCEL_MAX_MB: 100;
+    readonly SOURCE_CODE_MAX_MB: 20;
     readonly MAX_FILE_SIZE: number;
     readonly MAX_IMAGE_SIZE: number;
     readonly MAX_TEXT_SIZE: number;

package/dist/processors/config/sizeLimits.js

@@ -13,28 +13,34 @@
 export const SIZE_LIMITS_MB = {
     /** Maximum image file size (10MB) */
     IMAGE_MAX_MB: 10,
-    /** Maximum PDF file size (20MB) */
-    PDF_MAX_MB: 20,
-    /** Maximum document file size (10MB) */
-    DOCUMENT_MAX_MB: 10,
-    /** Maximum Word document size (10MB) */
-    WORD_MAX_MB: 10,
-    /** Maximum text file size (5MB) */
-    TEXT_MAX_MB: 5,
-    /** Maximum CSV file size (10MB) */
-    CSV_MAX_MB: 10,
-    /** Maximum Excel file size (10MB) */
-    EXCEL_MAX_MB: 10,
-    /** Maximum source code file size (5MB) */
-    SOURCE_CODE_MAX_MB: 5,
-    /** Maximum JSON file size (5MB) */
-    JSON_MAX_MB: 5,
-    /** Maximum YAML file size (5MB) */
-    YAML_MAX_MB: 5,
-    /** Maximum XML file size (5MB) */
-    XML_MAX_MB: 5,
-    /** Maximum general file size (50MB) */
-    GENERAL_MAX_MB: 50,
+    /** Maximum PDF file size (100MB — enterprise reports, long research papers) */
+    PDF_MAX_MB: 100,
+    /** Maximum document file size (100MB — enterprise docs with embedded images) */
+    DOCUMENT_MAX_MB: 100,
+    /** Maximum Word document size (100MB — PRDs, specs with embedded media) */
+    WORD_MAX_MB: 100,
+    /** Maximum text file size (50MB — large log files, build output) */
+    TEXT_MAX_MB: 50,
+    /** Maximum CSV file size (50MB — large data exports) */
+    CSV_MAX_MB: 50,
+    /** Maximum Excel file size (100MB — enterprise spreadsheets, data exports) */
+    EXCEL_MAX_MB: 100,
+    /** Maximum source code file size (20MB — large codebases) */
+    SOURCE_CODE_MAX_MB: 20,
+    /** Maximum JSON file size (20MB — large API response dumps) */
+    JSON_MAX_MB: 20,
+    /** Maximum YAML file size (20MB — large config files) */
+    YAML_MAX_MB: 20,
+    /** Maximum XML file size (20MB — large config/data files) */
+    XML_MAX_MB: 20,
+    /** Maximum video file size (500MB — long meeting recordings, screen captures) */
+    VIDEO_MAX_MB: 500,
+    /** Maximum audio file size (500MB — long meeting recordings, call recordings) */
+    AUDIO_MAX_MB: 500,
+    /** Maximum archive file size (200MB — large project archives) */
+    ARCHIVE_MAX_MB: 200,
+    /** Maximum general file size (200MB — matches Curator's memory-safety cap) */
+    GENERAL_MAX_MB: 200,
 };
 // =============================================================================
 // FILE SIZE LIMITS (in bytes)
@@ -45,28 +51,34 @@ export const SIZE_LIMITS_MB = {
 export const SIZE_LIMITS_BYTES = {
     /** Maximum image file size (10MB) */
     IMAGE_MAX: 10 * 1024 * 1024,
-    /** Maximum PDF file size (20MB) */
-    PDF_MAX: 20 * 1024 * 1024,
-    /** Maximum document file size (10MB) */
-    DOCUMENT_MAX: 10 * 1024 * 1024,
-    /** Maximum Word document size (10MB) */
-    WORD_MAX: 10 * 1024 * 1024,
-    /** Maximum text file size (5MB) */
-    TEXT_MAX: 5 * 1024 * 1024,
-    /** Maximum CSV file size (10MB) */
-    CSV_MAX: 10 * 1024 * 1024,
-    /** Maximum Excel file size (10MB) */
-    EXCEL_MAX: 10 * 1024 * 1024,
-    /** Maximum source code file size (5MB) */
-    SOURCE_CODE_MAX: 5 * 1024 * 1024,
-    /** Maximum JSON file size (5MB) */
-    JSON_MAX: 5 * 1024 * 1024,
-    /** Maximum YAML file size (5MB) */
-    YAML_MAX: 5 * 1024 * 1024,
-    /** Maximum XML file size (5MB) */
-    XML_MAX: 5 * 1024 * 1024,
-    /** Maximum general file size (50MB) */
-    GENERAL_MAX: 50 * 1024 * 1024,
+    /** Maximum PDF file size (100MB) */
+    PDF_MAX: 100 * 1024 * 1024,
+    /** Maximum document file size (100MB) */
+    DOCUMENT_MAX: 100 * 1024 * 1024,
+    /** Maximum Word document size (100MB) */
+    WORD_MAX: 100 * 1024 * 1024,
+    /** Maximum text file size (50MB) */
+    TEXT_MAX: 50 * 1024 * 1024,
+    /** Maximum CSV file size (50MB) */
+    CSV_MAX: 50 * 1024 * 1024,
+    /** Maximum Excel file size (100MB) */
+    EXCEL_MAX: 100 * 1024 * 1024,
+    /** Maximum source code file size (20MB) */
+    SOURCE_CODE_MAX: 20 * 1024 * 1024,
+    /** Maximum JSON file size (20MB) */
+    JSON_MAX: 20 * 1024 * 1024,
+    /** Maximum YAML file size (20MB) */
+    YAML_MAX: 20 * 1024 * 1024,
+    /** Maximum XML file size (20MB) */
+    XML_MAX: 20 * 1024 * 1024,
+    /** Maximum video file size (500MB) */
+    VIDEO_MAX: 500 * 1024 * 1024,
+    /** Maximum audio file size (500MB) */
+    AUDIO_MAX: 500 * 1024 * 1024,
+    /** Maximum archive file size (200MB) */
+    ARCHIVE_MAX: 200 * 1024 * 1024,
+    /** Maximum general file size (200MB) */
+    GENERAL_MAX: 200 * 1024 * 1024,
 };
 // =============================================================================
 // PROCESSING LIMITS

package/dist/processors/document/ExcelProcessor.d.ts

@@ -123,6 +123,20 @@ export declare class ExcelProcessor extends BaseFileProcessor<ProcessedExcel> {
      * @returns Concatenated plain text
      */
     private extractRichText;
+    /**
+     * Extract a specific range from a spreadsheet.
+     *
+     * Called by the `extract_file_content` tool for targeted data access.
+     * Returns TSV-formatted text for the specified sheet, row range, and columns.
+     *
+     * @param buffer - Excel file buffer
+     * @param sheet - Sheet name or 0-based index (default: first sheet)
+     * @param rowStart - Starting row (1-indexed, default: 1)
+     * @param rowEnd - Ending row (1-indexed, default: all rows)
+     * @param columns - Specific column letters to include (e.g., ["A", "B", "D"])
+     * @returns TSV-formatted string with the extracted data
+     */
+    extractSheetRange(buffer: Buffer, sheet?: string | number, rowStart?: number, rowEnd?: number, columns?: string[]): Promise<string>;
 }
 /**
  * Singleton Excel processor instance.

package/dist/processors/document/ExcelProcessor.js

@@ -35,7 +35,8 @@
  * }
  * ```
  */
-import { Workbook } from "exceljs";
+import ExcelJS from "exceljs";
+const { Workbook } = ExcelJS;
 import { BaseFileProcessor } from "../base/BaseFileProcessor.js";
 import { SIZE_LIMITS } from "../config/index.js";
 import { FileErrorCode } from "../errors/index.js";
@@ -385,6 +386,76 @@ export class ExcelProcessor extends BaseFileProcessor {
         })
             .join("");
     }
+    // ===========================================================================
+    // TARGETED EXTRACTION API
+    // ===========================================================================
+    /**
+     * Extract a specific range from a spreadsheet.
+     *
+     * Called by the `extract_file_content` tool for targeted data access.
+     * Returns TSV-formatted text for the specified sheet, row range, and columns.
+     *
+     * @param buffer - Excel file buffer
+     * @param sheet - Sheet name or 0-based index (default: first sheet)
+     * @param rowStart - Starting row (1-indexed, default: 1)
+     * @param rowEnd - Ending row (1-indexed, default: all rows)
+     * @param columns - Specific column letters to include (e.g., ["A", "B", "D"])
+     * @returns TSV-formatted string with the extracted data
+     */
+    async extractSheetRange(buffer, sheet, rowStart = 1, rowEnd, columns) {
+        const workbook = await this.parseWorkbook(buffer);
+        // Resolve the target worksheet
+        let worksheet;
+        if (typeof sheet === "number") {
+            // exceljs worksheets are 1-indexed
+            worksheet = workbook.worksheets[sheet];
+        }
+        else if (typeof sheet === "string") {
+            worksheet = workbook.getWorksheet(sheet);
+        }
+        else {
+            worksheet = workbook.worksheets[0];
+        }
+        if (!worksheet) {
+            const sheetNames = workbook.worksheets.map((ws) => ws.name).join(", ");
+            return `Sheet not found. Available sheets: ${sheetNames}`;
+        }
+        // Convert column letters to 1-based column indices if specified
+        const columnIndices = columns?.map((col) => {
+            let index = 0;
+            for (let i = 0; i < col.length; i++) {
+                index = index * 26 + col.toUpperCase().charCodeAt(i) - 64;
+            }
+            return index;
+        });
+        const lines = [];
+        lines.push(`## Sheet: ${worksheet.name}`);
+        const actualRowEnd = rowEnd ?? worksheet.rowCount;
+        let rowCount = 0;
+        worksheet.eachRow({ includeEmpty: false }, (row, rowNumber) => {
+            if (rowNumber < rowStart || rowNumber > actualRowEnd) {
+                return;
+            }
+            rowCount++;
+            const values = [];
+            row.eachCell({ includeEmpty: true }, (cell, colNumber) => {
+                if (columnIndices && !columnIndices.includes(colNumber)) {
+                    return;
+                }
+                const val = this.getCellValue(cell.value);
+                values.push(val === null ? "" : String(val));
+            });
+            // Add row number prefix for easy reference
+            lines.push(`${rowNumber}\t${values.join("\t")}`);
+        });
+        if (rowCount === 0) {
+            lines.push(`(No data in rows ${rowStart}-${actualRowEnd})`);
+        }
+        else {
+            lines.push(`\n(${rowCount} rows, range ${rowStart}-${actualRowEnd})`);
+        }
+        return lines.join("\n");
+    }
 }
 // =============================================================================
 // SINGLETON INSTANCE

package/dist/processors/document/PptxProcessor.d.ts

@@ -0,0 +1,63 @@
+/**
+ * PowerPoint (PPTX) Processing Utility
+ *
+ * Extracts text content from PowerPoint (.pptx) files by treating them
+ * as ZIP archives and parsing the slide XML files within.
+ *
+ * PPTX files are ZIP archives containing:
+ * - ppt/slides/slide1.xml, slide2.xml, ... — slide content
+ * - ppt/slideMasters/ — master slide templates
+ * - ppt/slideLayouts/ — slide layout definitions
+ *
+ * Text is extracted from `<a:t>` elements in the slide XML files.
+ * Slides are sorted by number and presented in reading order.
+ *
+ * Uses `adm-zip` (already a project dependency) for ZIP extraction.
+ *
+ * @module processors/document/PptxProcessor
+ *
+ * @example
+ * ```typescript
+ * import { PptxProcessor } from "./PptxProcessor.js";
+ *
+ * const text = await PptxProcessor.extractText(buffer);
+ * if (text) {
+ *   console.log("Extracted text:", text);
+ * }
+ * ```
+ */
+/**
+ * Static utility class for extracting text from PPTX files.
+ *
+ * Designed as a static class (not extending BaseFileProcessor) because
+ * PPTX processing is straightforward ZIP+XML extraction and does not
+ * need the full download/validate/process pipeline of BaseFileProcessor.
+ */
+export declare class PptxProcessor {
+    /**
+     * Extract all text content from a PPTX buffer.
+     *
+     * @param content - Raw PPTX file buffer
+     * @returns Formatted text content with slide headers, or null if no text found
+     * @throws Error if the buffer is not a valid ZIP/PPTX file
+     */
+    static extractText(content: Buffer): Promise<string | null>;
+    /**
+     * Extract text strings from a slide XML document.
+     * Finds all `<a:t>` elements and returns their text content.
+     *
+     * @param xml - Raw XML string from a slide file
+     * @returns Array of text strings found in the slide
+     */
+    private static extractTextFromXml;
+    /**
+     * Extract text from specific slides in a PPTX file.
+     *
+     * Called by the `extract_file_content` tool for targeted slide access.
+     *
+     * @param content - Raw PPTX file buffer
+     * @param slideNumbers - Array of 1-indexed slide numbers to extract
+     * @returns Formatted text from the requested slides
+     */
+    static extractSlides(content: Buffer, slideNumbers: number[]): Promise<string>;
+}

package/dist/processors/document/PptxProcessor.js

@@ -0,0 +1,157 @@
+/**
+ * PowerPoint (PPTX) Processing Utility
+ *
+ * Extracts text content from PowerPoint (.pptx) files by treating them
+ * as ZIP archives and parsing the slide XML files within.
+ *
+ * PPTX files are ZIP archives containing:
+ * - ppt/slides/slide1.xml, slide2.xml, ... — slide content
+ * - ppt/slideMasters/ — master slide templates
+ * - ppt/slideLayouts/ — slide layout definitions
+ *
+ * Text is extracted from `<a:t>` elements in the slide XML files.
+ * Slides are sorted by number and presented in reading order.
+ *
+ * Uses `adm-zip` (already a project dependency) for ZIP extraction.
+ *
+ * @module processors/document/PptxProcessor
+ *
+ * @example
+ * ```typescript
+ * import { PptxProcessor } from "./PptxProcessor.js";
+ *
+ * const text = await PptxProcessor.extractText(buffer);
+ * if (text) {
+ *   console.log("Extracted text:", text);
+ * }
+ * ```
+ */
+import AdmZip from "adm-zip";
+/**
+ * Regex to match text content within PowerPoint XML `<a:t>` elements.
+ * These elements contain the actual visible text on slides.
+ */
+const TEXT_ELEMENT_REGEX = /<a:t[^>]*>([\s\S]*?)<\/a:t>/g;
+/**
+ * Regex to match slide filenames and extract slide number.
+ * Matches entries like "ppt/slides/slide1.xml", "ppt/slides/slide12.xml".
+ */
+const SLIDE_ENTRY_REGEX = /^ppt\/slides\/slide(\d+)\.xml$/;
+/**
+ * Static utility class for extracting text from PPTX files.
+ *
+ * Designed as a static class (not extending BaseFileProcessor) because
+ * PPTX processing is straightforward ZIP+XML extraction and does not
+ * need the full download/validate/process pipeline of BaseFileProcessor.
+ */
+export class PptxProcessor {
+    /**
+     * Extract all text content from a PPTX buffer.
+     *
+     * @param content - Raw PPTX file buffer
+     * @returns Formatted text content with slide headers, or null if no text found
+     * @throws Error if the buffer is not a valid ZIP/PPTX file
+     */
+    static async extractText(content) {
+        const zip = new AdmZip(content);
+        const entries = zip.getEntries();
+        // Collect slide entries with their slide numbers for sorting
+        const slides = [];
+        for (const entry of entries) {
+            const match = entry.entryName.match(SLIDE_ENTRY_REGEX);
+            if (match) {
+                const slideNumber = parseInt(match[1], 10);
+                const xmlContent = entry.getData().toString("utf-8");
+                slides.push({ slideNumber, xml: xmlContent });
+            }
+        }
+        // Sort slides by number (slide1, slide2, ...)
+        slides.sort((a, b) => a.slideNumber - b.slideNumber);
+        if (slides.length === 0) {
+            return null;
+        }
+        const parts = [];
+        parts.push(`Presentation: ${slides.length} slide(s)\n`);
+        for (const slide of slides) {
+            const texts = PptxProcessor.extractTextFromXml(slide.xml);
+            if (texts.length > 0) {
+                parts.push(`### Slide ${slide.slideNumber}`);
+                parts.push(texts.join("\n"));
+                parts.push(""); // blank line between slides
+            }
+        }
+        const result = parts.join("\n").trim();
+        return result || null;
+    }
+    /**
+     * Extract text strings from a slide XML document.
+     * Finds all `<a:t>` elements and returns their text content.
+     *
+     * @param xml - Raw XML string from a slide file
+     * @returns Array of text strings found in the slide
+     */
+    static extractTextFromXml(xml) {
+        const texts = [];
+        // Reset regex state for re-entrant usage
+        TEXT_ELEMENT_REGEX.lastIndex = 0;
+        for (let match = TEXT_ELEMENT_REGEX.exec(xml); match !== null; match = TEXT_ELEMENT_REGEX.exec(xml)) {
+            const text = match[1].trim();
+            if (text) {
+                texts.push(text);
+            }
+        }
+        return texts;
+    }
+    // ===========================================================================
+    // TARGETED EXTRACTION API
+    // ===========================================================================
+    /**
+     * Extract text from specific slides in a PPTX file.
+     *
+     * Called by the `extract_file_content` tool for targeted slide access.
+     *
+     * @param content - Raw PPTX file buffer
+     * @param slideNumbers - Array of 1-indexed slide numbers to extract
+     * @returns Formatted text from the requested slides
+     */
+    static async extractSlides(content, slideNumbers) {
+        const zip = new AdmZip(content);
+        const entries = zip.getEntries();
+        // Collect all slides
+        const slides = [];
+        for (const entry of entries) {
+            const match = entry.entryName.match(SLIDE_ENTRY_REGEX);
+            if (match) {
+                const slideNumber = parseInt(match[1], 10);
+                if (slideNumbers.includes(slideNumber)) {
+                    const xmlContent = entry.getData().toString("utf-8");
+                    slides.push({ slideNumber, xml: xmlContent });
+                }
+            }
+        }
+        slides.sort((a, b) => a.slideNumber - b.slideNumber);
+        if (slides.length === 0) {
+            // List total slides to help the LLM
+            let totalSlides = 0;
+            for (const entry of entries) {
+                if (SLIDE_ENTRY_REGEX.test(entry.entryName)) {
+                    totalSlides++;
+                }
+            }
+            return `Slides ${slideNumbers.join(", ")} not found. This presentation has ${totalSlides} slide(s).`;
+        }
+        const parts = [];
+        for (const slide of slides) {
+            const texts = PptxProcessor.extractTextFromXml(slide.xml);
+            parts.push(`### Slide ${slide.slideNumber}`);
+            if (texts.length > 0) {
+                parts.push(texts.join("\n"));
+            }
+            else {
+                parts.push("(No text content on this slide)");
+            }
+            parts.push("");
+        }
+        return parts.join("\n").trim();
+    }
+}

package/dist/processors/document/index.d.ts

@@ -50,5 +50,6 @@
  */
 export { isWordFile, type ProcessedWord, processWord, validateWordSize, WordProcessor, wordProcessor, } from "./WordProcessor.js";
 export { ExcelProcessor, type ExcelWorksheet, excelProcessor, getExcelMaxRows, getExcelMaxSheets, getExcelMaxSizeMB, isExcelFile, type ProcessedExcel, processExcel, validateExcelSize, } from "./ExcelProcessor.js";
+export { PptxProcessor, } from "./PptxProcessor.js";
 export { isRtfFile, type ProcessedRtf, processRtf, RtfProcessor, rtfProcessor, validateRtfSize, } from "./RtfProcessor.js";
 export { getOpenDocumentMaxSizeMB, isOpenDocumentFile, OpenDocumentProcessor, openDocumentProcessor, type ProcessedOpenDocument, processOpenDocument, validateOpenDocumentSize, } from "./OpenDocumentProcessor.js";

package/dist/processors/document/index.js

@@ -69,6 +69,12 @@ excelProcessor,
 // Helper functions
 getExcelMaxRows, getExcelMaxSheets, getExcelMaxSizeMB, isExcelFile, processExcel, validateExcelSize, } from "./ExcelProcessor.js";
 // =============================================================================
+// PPTX PROCESSOR
+// =============================================================================
+export { 
+// Class
+PptxProcessor, } from "./PptxProcessor.js";
+// =============================================================================
 // RTF PROCESSOR
 // =============================================================================
 export { 

package/dist/processors/errors/FileErrorCode.d.ts

@@ -67,7 +67,7 @@ export declare enum FileErrorCode {
 /**
  * Error message template with user-friendly messaging and retry information.
  */
-export interface ErrorMessageTemplate {
+export type ErrorMessageTemplate = {
     /** Technical error message */
     message: string;
     /** User-friendly error message */
@@ -76,7 +76,7 @@ export interface ErrorMessageTemplate {
     suggestedAction: string;
     /** Whether this error is potentially retryable */
     retryable: boolean;
-}
+};
 /**
  * Error messages map with technical and user-friendly messaging for each error code.
  * All messages are designed to be clear, actionable, and free of technical jargon.

package/dist/processors/errors/errorHelpers.d.ts

@@ -13,7 +13,7 @@ export type { FileProcessingError };
 /**
  * Summary of file processing operations.
  */
-export interface FileProcessingSummary {
+export type FileProcessingSummary = {
     /** Total number of files attempted */
     totalFiles: number;
     /** Successfully processed files */
@@ -38,7 +38,7 @@ export interface FileProcessingSummary {
         filename: string;
         message: string;
     }>;
-}
+};
 /**
  * Create a structured file processing error with user-friendly messaging.
  *