@apify/actors-mcp-server 0.4.9 → 0.4.11

package/dist/utils/generic.d.ts

@@ -1,3 +1,14 @@
+/**
+ * Parses a comma-separated string into an array of trimmed strings.
+ * Empty strings are filtered out after trimming.
+ *
+ * @param input - The comma-separated string to parse. If undefined, returns an empty array.
+ * @returns An array of trimmed, non-empty strings.
+ * @example
+ * parseCommaSeparatedList("a, b, c"); // ["a", "b", "c"]
+ * parseCommaSeparatedList("a, , b"); // ["a", "b"]
+ */
+export declare function parseCommaSeparatedList(input?: string): string[];
 /**
  * Recursively gets the value in a nested object for each key in the keys array.
  * Each key can be a dot-separated path (e.g. 'a.b.c').
@@ -8,16 +19,11 @@
  * const value = getValuesByDotKeys(obj, ['a.b.c', 'a.b.d', 'nested']);
  * value; // { 'a.b.c': 42, 'a.b.d': undefined, 'nested': { d: 100 } }
  */
+export declare function getValuesByDotKeys(obj: Record<string, unknown>, keys: string[]): Record<string, unknown>;
 /**
- * Parses a comma-separated string into an array of trimmed strings.
- * Empty strings are filtered out after trimming.
+ * Validates whether a given string is a well-formed URL.
  *
- * @param input - The comma-separated string to parse. If undefined, returns an empty array.
- * @returns An array of trimmed, non-empty strings.
- * @example
- * parseCommaSeparatedList("a, b, c"); // ["a", "b", "c"]
- * parseCommaSeparatedList("a, , b"); // ["a", "b"]
+ * Allows only valid HTTP or HTTPS URLs.
  */
-export declare function parseCommaSeparatedList(input?: string): string[];
-export declare function getValuesByDotKeys(obj: Record<string, unknown>, keys: string[]): Record<string, unknown>;
+export declare function isValidHttpUrl(urlString: string): boolean;
 //# sourceMappingURL=generic.d.ts.map

package/dist/utils/generic.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"generic.d.ts","sourceRoot":"","sources":["../../src/utils/generic.ts"],"names":[],"mappings":"AAAA;;;;;;;;;GASG;AACH;;;;;;;;;GASG;AACH,wBAAgB,uBAAuB,CAAC,KAAK,CAAC,EAAE,MAAM,GAAG,MAAM,EAAE,CAKhE;AAED,wBAAgB,kBAAkB,CAAC,GAAG,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,EAAE,IAAI,EAAE,MAAM,EAAE,GAAG,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAqBxG"}
+{"version":3,"file":"generic.d.ts","sourceRoot":"","sources":["../../src/utils/generic.ts"],"names":[],"mappings":"AAAA;;;;;;;;;GASG;AACH,wBAAgB,uBAAuB,CAAC,KAAK,CAAC,EAAE,MAAM,GAAG,MAAM,EAAE,CAKhE;AAED;;;;;;;;;GASG;AACH,wBAAgB,kBAAkB,CAAC,GAAG,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,EAAE,IAAI,EAAE,MAAM,EAAE,GAAG,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAqBxG;AAED;;;;GAIG;AACH,wBAAgB,cAAc,CAAC,SAAS,EAAE,MAAM,GAAG,OAAO,CAWzD"}

package/dist/utils/generic.js

@@ -1,13 +1,3 @@
-/**
- * Recursively gets the value in a nested object for each key in the keys array.
- * Each key can be a dot-separated path (e.g. 'a.b.c').
- * Returns an object mapping each key to its resolved value (or undefined if not found).
- *
- * @example
- * const obj = { a: { b: { c: 42 } }, nested: { d: 100 } };
- * const value = getValuesByDotKeys(obj, ['a.b.c', 'a.b.d', 'nested']);
- * value; // { 'a.b.c': 42, 'a.b.d': undefined, 'nested': { d: 100 } }
- */
 /**
  * Parses a comma-separated string into an array of trimmed strings.
  * Empty strings are filtered out after trimming.
@@ -24,6 +14,16 @@ export function parseCommaSeparatedList(input) {
     }
     return input.split(',').map((s) => s.trim()).filter((s) => s.length > 0);
 }
+/**
+ * Recursively gets the value in a nested object for each key in the keys array.
+ * Each key can be a dot-separated path (e.g. 'a.b.c').
+ * Returns an object mapping each key to its resolved value (or undefined if not found).
+ *
+ * @example
+ * const obj = { a: { b: { c: 42 } }, nested: { d: 100 } };
+ * const value = getValuesByDotKeys(obj, ['a.b.c', 'a.b.d', 'nested']);
+ * value; // { 'a.b.c': 42, 'a.b.d': undefined, 'nested': { d: 100 } }
+ */
 export function getValuesByDotKeys(obj, keys) {
     const result = {};
     for (const key of keys) {
@@ -45,4 +45,22 @@ export function getValuesByDotKeys(obj, keys) {
     }
     return result;
 }
+/**
+ * Validates whether a given string is a well-formed URL.
+ *
+ * Allows only valid HTTP or HTTPS URLs.
+ */
+export function isValidHttpUrl(urlString) {
+    if (!urlString.startsWith('http://') && !urlString.startsWith('https://')) {
+        return false;
+    }
+    try {
+        /* eslint-disable no-new */
+        new URL(urlString);
+        return true;
+    }
+    catch {
+        return false;
+    }
+}
 //# sourceMappingURL=generic.js.map

package/dist/utils/generic.js.map

@@ -1 +1 @@
-{"version":3,"file":"generic.js","sourceRoot":"","sources":["../../src/utils/generic.ts"],"names":[],"mappings":"AAAA;;;;;;;;;GASG;AACH;;;;;;;;;GASG;AACH,MAAM,UAAU,uBAAuB,CAAC,KAAc;IAClD,IAAI,CAAC,KAAK,EAAE,CAAC;QACT,OAAO,EAAE,CAAC;IACd,CAAC;IACD,OAAO,KAAK,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;AAC7E,CAAC;AAED,MAAM,UAAU,kBAAkB,CAAC,GAA4B,EAAE,IAAc;IAC3E,MAAM,MAAM,GAA4B,EAAE,CAAC;IAC3C,KAAK,MAAM,GAAG,IAAI,IAAI,EAAE,CAAC;QACrB,MAAM,IAAI,GAAG,GAAG,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC;QAC5B,IAAI,OAAO,GAAY,GAAG,CAAC;QAC3B,KAAK,MAAM,OAAO,IAAI,IAAI,EAAE,CAAC;YACzB,IACI,OAAO,KAAK,IAAI;mBACb,OAAO,OAAO,KAAK,QAAQ;mBAC3B,MAAM,CAAC,SAAS,CAAC,cAAc,CAAC,IAAI,CAAC,OAAO,EAAE,OAAO,CAAC,EAC3D,CAAC;gBACC,qDAAqD;gBACrD,OAAO,GAAI,OAAmC,CAAC,OAAO,CAAC,CAAC;YAC5D,CAAC;iBAAM,CAAC;gBACJ,OAAO,GAAG,SAAS,CAAC;gBACpB,MAAM;YACV,CAAC;QACL,CAAC;QACD,MAAM,CAAC,GAAG,CAAC,GAAG,OAAO,CAAC;IAC1B,CAAC;IACD,OAAO,MAAM,CAAC;AAClB,CAAC"}
+{"version":3,"file":"generic.js","sourceRoot":"","sources":["../../src/utils/generic.ts"],"names":[],"mappings":"AAAA;;;;;;;;;GASG;AACH,MAAM,UAAU,uBAAuB,CAAC,KAAc;IAClD,IAAI,CAAC,KAAK,EAAE,CAAC;QACT,OAAO,EAAE,CAAC;IACd,CAAC;IACD,OAAO,KAAK,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;AAC7E,CAAC;AAED;;;;;;;;;GASG;AACH,MAAM,UAAU,kBAAkB,CAAC,GAA4B,EAAE,IAAc;IAC3E,MAAM,MAAM,GAA4B,EAAE,CAAC;IAC3C,KAAK,MAAM,GAAG,IAAI,IAAI,EAAE,CAAC;QACrB,MAAM,IAAI,GAAG,GAAG,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC;QAC5B,IAAI,OAAO,GAAY,GAAG,CAAC;QAC3B,KAAK,MAAM,OAAO,IAAI,IAAI,EAAE,CAAC;YACzB,IACI,OAAO,KAAK,IAAI;mBACb,OAAO,OAAO,KAAK,QAAQ;mBAC3B,MAAM,CAAC,SAAS,CAAC,cAAc,CAAC,IAAI,CAAC,OAAO,EAAE,OAAO,CAAC,EAC3D,CAAC;gBACC,qDAAqD;gBACrD,OAAO,GAAI,OAAmC,CAAC,OAAO,CAAC,CAAC;YAC5D,CAAC;iBAAM,CAAC;gBACJ,OAAO,GAAG,SAAS,CAAC;gBACpB,MAAM;YACV,CAAC;QACL,CAAC;QACD,MAAM,CAAC,GAAG,CAAC,GAAG,OAAO,CAAC;IAC1B,CAAC;IACD,OAAO,MAAM,CAAC;AAClB,CAAC;AAED;;;;GAIG;AACH,MAAM,UAAU,cAAc,CAAC,SAAiB;IAC5C,IAAI,CAAC,SAAS,CAAC,UAAU,CAAC,SAAS,CAAC,IAAI,CAAC,SAAS,CAAC,UAAU,CAAC,UAAU,CAAC,EAAE,CAAC;QACxE,OAAO,KAAK,CAAC;IACjB,CAAC;IACD,IAAI,CAAC;QACD,2BAA2B;QAC3B,IAAI,GAAG,CAAC,SAAS,CAAC,CAAC;QACnB,OAAO,IAAI,CAAC;IAChB,CAAC;IAAC,MAAM,CAAC;QACL,OAAO,KAAK,CAAC;IACjB,CAAC;AACL,CAAC"}

package/dist/utils/html.d.ts

@@ -0,0 +1,11 @@
+/**
+ * Strips HTML and keeps only the structure.
+ *
+ * Removes styles, scripts, and other non-content elements.
+ * Collapses whitespace and trims the result.
+ * Keeps only href, src, alt, id, class, title, name, data-* attributes.
+ * Removes HTML comments and spaces between tags.
+ * Removes base64 encoded images.
+ */
+export declare function stripHtml(html: string): string;
+//# sourceMappingURL=html.d.ts.map

package/dist/utils/html.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"html.d.ts","sourceRoot":"","sources":["../../src/utils/html.ts"],"names":[],"mappings":"AAWA;;;;;;;;GAQG;AACH,wBAAgB,SAAS,CAAC,IAAI,EAAE,MAAM,GAAG,MAAM,CAuC9C"}

package/dist/utils/html.js

@@ -0,0 +1,48 @@
+import * as cheerio from 'cheerio';
+/**
+ * Strips HTML and keeps only the structure.
+ *
+ * Removes styles, scripts, and other non-content elements.
+ * Collapses whitespace and trims the result.
+ * Keeps only href, src, alt, id, class, title, name, data-* attributes.
+ * Removes HTML comments and spaces between tags.
+ * Removes base64 encoded images.
+ */
+export function stripHtml(html) {
+    const $ = cheerio.load(html);
+    // Remove all attributes except href (only on a), src, alt, id, class, title, name, data-*
+    const allowedAttrs = ['href', 'src', 'alt', 'id', 'class', 'title', 'name'];
+    $('*').each((_, element) => {
+        const { attribs } = element;
+        if (attribs) {
+            Object.keys(attribs).forEach((attr) => {
+                if (attr === 'href' && element.tagName !== 'a') {
+                    $(element).removeAttr(attr);
+                }
+                else if (!allowedAttrs.includes(attr) && !attr.startsWith('data-')) {
+                    $(element).removeAttr(attr);
+                }
+            });
+        }
+    });
+    // Remove <style>, <script>, <noscript>, <iframe>, <svg>, <canvas>, <math> tags and their content
+    $('style, script, noscript, iframe, svg, canvas, math').remove();
+    // Remove HTML comments
+    $('*').contents().filter((_, element) => element.type === 'comment').remove();
+    // Remove base64 encoded images
+    $('img[src^="data:image/"]').remove();
+    let result;
+    if (html.trim() === '') {
+        result = '';
+    }
+    else if (html.includes('<html')) {
+        result = $.html();
+    }
+    else {
+        result = $('body').html() || '';
+    }
+    // Collapse multiple spaces into one, remove spaces between tags, and trim
+    result = result.replace(/\s+/g, ' ').replace(/>\s+</g, '><').trim();
+    return result;
+}
+//# sourceMappingURL=html.js.map

package/dist/utils/html.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"html.js","sourceRoot":"","sources":["../../src/utils/html.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,OAAO,MAAM,SAAS,CAAC;AAWnC;;;;;;;;GAQG;AACH,MAAM,UAAU,SAAS,CAAC,IAAY;IAClC,MAAM,CAAC,GAAG,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IAE7B,0FAA0F;IAC1F,MAAM,YAAY,GAAG,CAAC,MAAM,EAAE,KAAK,EAAE,KAAK,EAAE,IAAI,EAAE,OAAO,EAAE,OAAO,EAAE,MAAM,CAAC,CAAC;IAC5E,CAAC,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,OAAO,EAAE,EAAE;QACvB,MAAM,EAAE,OAAO,EAAE,GAAI,OAA8B,CAAC;QACpD,IAAI,OAAO,EAAE,CAAC;YACV,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,OAAO,CAAC,CAAC,IAAI,EAAE,EAAE;gBAClC,IAAI,IAAI,KAAK,MAAM,IAAK,OAA8B,CAAC,OAAO,KAAK,GAAG,EAAE,CAAC;oBACrE,CAAC,CAAC,OAAO,CAAC,CAAC,UAAU,CAAC,IAAI,CAAC,CAAC;gBAChC,CAAC;qBAAM,IAAI,CAAC,YAAY,CAAC,QAAQ,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,UAAU,CAAC,OAAO,CAAC,EAAE,CAAC;oBACnE,CAAC,CAAC,OAAO,CAAC,CAAC,UAAU,CAAC,IAAI,CAAC,CAAC;gBAChC,CAAC;YACL,CAAC,CAAC,CAAC;QACP,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,iGAAiG;IACjG,CAAC,CAAC,oDAAoD,CAAC,CAAC,MAAM,EAAE,CAAC;IAEjE,uBAAuB;IACvB,CAAC,CAAC,GAAG,CAAC,CAAC,QAAQ,EAAE,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,OAAO,EAAE,EAAE,CAAE,OAAoB,CAAC,IAAI,KAAK,SAAS,CAAC,CAAC,MAAM,EAAE,CAAC;IAE5F,+BAA+B;IAC/B,CAAC,CAAC,yBAAyB,CAAC,CAAC,MAAM,EAAE,CAAC;IAEtC,IAAI,MAAM,CAAC;IACX,IAAI,IAAI,CAAC,IAAI,EAAE,KAAK,EAAE,EAAE,CAAC;QACrB,MAAM,GAAG,EAAE,CAAC;IAChB,CAAC;SAAM,IAAI,IAAI,CAAC,QAAQ,CAAC,OAAO,CAAC,EAAE,CAAC;QAChC,MAAM,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC;IACtB,CAAC;SAAM,CAAC;QACJ,MAAM,GAAG,CAAC,CAAC,MAAM,CAAC,CAAC,IAAI,EAAE,IAAI,EAAE,CAAC;IACpC,CAAC;IAED,0EAA0E;IAC1E,MAAM,GAAG,MAAM,CAAC,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC,OAAO,CAAC,QAAQ,EAAE,IAAI,CAAC,CAAC,IAAI,EAAE,CAAC;IACpE,OAAO,MAAM,CAAC;AAClB,CAAC"}

package/dist/utils/mcp.d.ts

@@ -0,0 +1,10 @@
+/**
+ * Helper to build a response for MCP from an array of text strings.
+ */
+export declare function buildMCPResponse(texts: string[]): {
+    content: {
+        type: string;
+        text: string;
+    }[];
+};
+//# sourceMappingURL=mcp.d.ts.map

package/dist/utils/mcp.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"mcp.d.ts","sourceRoot":"","sources":["../../src/utils/mcp.ts"],"names":[],"mappings":"AAAA;;GAEG;AACH,wBAAgB,gBAAgB,CAAC,KAAK,EAAE,MAAM,EAAE;;;;;EAI/C"}

package/dist/utils/mcp.js

@@ -0,0 +1,9 @@
+/**
+ * Helper to build a response for MCP from an array of text strings.
+ */
+export function buildMCPResponse(texts) {
+    return {
+        content: texts.map((text) => ({ type: 'text', text })),
+    };
+}
+//# sourceMappingURL=mcp.js.map

package/dist/utils/mcp.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"mcp.js","sourceRoot":"","sources":["../../src/utils/mcp.ts"],"names":[],"mappings":"AAAA;;GAEG;AACH,MAAM,UAAU,gBAAgB,CAAC,KAAe;IAC5C,OAAO;QACH,OAAO,EAAE,KAAK,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,CAAC,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,CAAC,CAAC;KACzD,CAAC;AACN,CAAC"}

package/package.json

@@ -1,8 +1,9 @@
 {
   "name": "@apify/actors-mcp-server",
-  "version": "0.4.9",
+  "version": "0.4.11",
   "type": "module",
   "description": "Apify MCP Server",
+  "mcpName": "com.apify/apify-mcp-server",
   "engines": {
     "node": ">=18.0.0"
   },
@@ -38,11 +39,13 @@
     "@apify/datastructures": "^2.0.3",
     "@apify/log": "^2.5.16",
     "@modelcontextprotocol/sdk": "^1.17.4",
+    "@types/cheerio": "^0.22.35",
     "@types/turndown": "^5.0.5",
     "ajv": "^8.17.1",
     "algoliasearch": "^5.31.0",
     "apify": "^3.4.2",
     "apify-client": "^2.12.6",
+    "cheerio": "^1.1.2",
     "express": "^4.21.2",
     "to-json-schema": "^0.2.5",
     "turndown": "^7.2.0",
@@ -71,6 +74,7 @@
     "start": "npm run start:dev",
     "start:prod": "node dist/main.js",
     "start:dev": "tsx src/main.ts",
+    "start:standby": "APIFY_META_ORIGIN=STANDBY npm run start",
     "lint": "eslint .",
     "lint:fix": "eslint . --fix",
     "build": "tsc -b src",