@sisu-ai/tool-extract-urls 2.0.0

package/README.md

@@ -0,0 +1,34 @@
+# @sisu-ai/tool-extract-urls
+
+Extract unique `http`/`https` URLs from text snippets. Small, deterministic, and zero I/O — great as a first pass before fetching, classifying, or summarizing pages.
+
+## Install
+```bash
+npm i @sisu-ai/tool-extract-urls
+```
+
+## Why it’s useful
+- Simple guardrail: avoids asking the model to spot links.
+- Deterministic: same inputs → same outputs.
+- Lightweight: no network calls, safe to run early in a pipeline.
+
+## Usage
+```ts
+import { Agent, SimpleTools } from '@sisu-ai/core';
+import { registerTools } from '@sisu-ai/mw-register-tools';
+import { toolCalling } from '@sisu-ai/mw-tool-calling';
+import { extractUrlsTool } from '@sisu-ai/tool-extract-urls';
+
+const app = new Agent()
+  .use(registerTools([extractUrlsTool]))
+  .use(toolCalling);
+
+// Prompt example: "Find links in: https://example.com and http://sisu.ai"
+```
+
+## What it returns
+- Array of unique URLs, e.g. `["https://example.com", "http://sisu.ai"]`.
+
+## Notes
+- The regex targets `http`/`https` URLs and ignores surrounding punctuation where possible.
+- Prefer pairing with a fetch tool (e.g., `@sisu-ai/tool-web-fetch`) for subsequent content analysis.

package/dist/index.d.ts

@@ -0,0 +1,18 @@
+import type { Tool } from '@sisu-ai/core';
+/**
+ * Extract all unique URLs from an array of strings.
+ *
+ * @param contents - Array of strings to scan for URLs.
+ * @returns Array of unique URLs found in the contents.
+ */
+export declare function extractUrls(contents: string[]): string[];
+/**
+ * Tool definition wrapping {@link extractUrls} for use with agents.
+ */
+export declare const extractUrlsTool: Tool<{
+    contents: string[];
+}>;
+declare const _default: Tool<{
+    contents: string[];
+}, unknown>[];
+export default _default;

package/dist/index.js

@@ -0,0 +1,25 @@
+import { z } from 'zod';
+/**
+ * Extract all unique URLs from an array of strings.
+ *
+ * @param contents - Array of strings to scan for URLs.
+ * @returns Array of unique URLs found in the contents.
+ */
+export function extractUrls(contents) {
+    const urlRe = /https?:\/\/[\S)\]"'>]+/gi;
+    const out = new Set();
+    for (const c of contents)
+        for (const m of c.matchAll(urlRe))
+            out.add(m[0]);
+    return Array.from(out);
+}
+/**
+ * Tool definition wrapping {@link extractUrls} for use with agents.
+ */
+export const extractUrlsTool = {
+    name: 'extractUrls',
+    description: 'Extract unique HTTP/HTTPS URLs from provided text snippets.',
+    schema: z.object({ contents: z.array(z.string()) }),
+    handler: async ({ contents }) => extractUrls(contents),
+};
+export default [extractUrlsTool];

package/package.json

@@ -0,0 +1,28 @@
+{
+  "name": "@sisu-ai/tool-extract-urls",
+  "version": "2.0.0",
+  "type": "module",
+  "main": "dist/index.js",
+  "types": "dist/index.d.ts",
+  "files": [
+    "dist"
+  ],
+  "scripts": {
+    "build": "tsc -b"
+  },
+  "dependencies": {
+    "zod": "^3.23.8"
+  },
+  "peerDependencies": {
+    "@sisu-ai/core": "0.3.0"
+  },
+  "repository": {
+    "type": "git",
+    "url": "https://github.com/finger-gun/sisu",
+    "directory": "packages/tools/extract-urls"
+  },
+  "homepage": "https://github.com/finger-gun/sisu#readme",
+  "bugs": {
+    "url": "https://github.com/finger-gun/sisu/issues"
+  }
+}