@webcrawl/mcp 1.3.10 → 1.4.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +1 -1
- package/dist/server.js +4 -4
- package/package.json +2 -2
package/README.md
CHANGED
|
@@ -46,7 +46,7 @@ Set `WEBCRAWL_MCP_SCHEMA_PROFILE=compact` (default when installed via `npx webcr
|
|
|
46
46
|
## Tools Provided
|
|
47
47
|
|
|
48
48
|
- `web_search` — Search the web or structured verticals (`shopping`, `flights`, `hotels`, `real-estate`, `business`, `security`, `academic`, `tech`, `news`).
|
|
49
|
-
- `
|
|
49
|
+
- `web_read` — Fetch a URL and return clean Markdown with target evidence.
|
|
50
50
|
- `web_parse` / `web_parse_status` — Parse PDFs (native + OCR), Word/Excel/PowerPoint files, YouTube videos, and audio/video into Markdown.
|
|
51
51
|
- `web_map` — Discover all URLs on a domain (`sitemap.xml`, `robots.txt`, and link graph).
|
|
52
52
|
- `web_crawl` — Crawl up to 25 pages starting from a URL.
|
package/dist/server.js
CHANGED
|
@@ -4,7 +4,7 @@ import { z } from "zod";
|
|
|
4
4
|
import { formatCredits, formatCrawl, formatMap, formatMonitor, formatParse, formatScrape, formatSearch, truncate, } from "./format.js";
|
|
5
5
|
const INSTRUCTIONS = `Webcrawl Anything gives you live web & target inspection access (zero-key local engine or webcrawl.app cloud).
|
|
6
6
|
- web_search: find pages or structured results. Set category for flights, hotels, real-estate, shopping, news, local business, security (CVEs), academic or tech. Results are localised to the user.
|
|
7
|
-
-
|
|
7
|
+
- web_read: read one URL as clean markdown with target evidence (websites, SPAs, GitHub, ArXiv, llms.txt).
|
|
8
8
|
- web_parse: read a PDF, Office file, YouTube video, image, or audio/video at a URL as markdown (OCR and transcripts included). For long files set background and poll web_parse_status.
|
|
9
9
|
- web_map: list the URLs on a site before deciding what to scrape.
|
|
10
10
|
- web_crawl: read several pages of one site (max 25).
|
|
@@ -68,11 +68,11 @@ export function createWebcrawlMcpServer(client, versionOrOpts = "0.1.0") {
|
|
|
68
68
|
});
|
|
69
69
|
return formatSearch(res) + formatCredits(res);
|
|
70
70
|
}));
|
|
71
|
-
server.registerTool("
|
|
72
|
-
title: "
|
|
71
|
+
server.registerTool("web_read", {
|
|
72
|
+
title: "Read a page",
|
|
73
73
|
description: compact
|
|
74
74
|
? "Fetch one URL as clean markdown."
|
|
75
|
-
: "Fetch one URL and return its main content as markdown.
|
|
75
|
+
: "Fetch one URL and return its main content as markdown.",
|
|
76
76
|
inputSchema: {
|
|
77
77
|
url: z.string().url(),
|
|
78
78
|
no_cache: desc(z.boolean().optional(), "Skip cached copy"),
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@webcrawl/mcp",
|
|
3
|
-
"version": "1.
|
|
3
|
+
"version": "1.4.0",
|
|
4
4
|
"mcpName": "io.github.brainfish-ai/webcrawl",
|
|
5
5
|
"description": "MCP server for Webcrawl Anything: search, scrape, map, crawl, parse and monitor the web from your agent. Runs locally with no API key, or on webcrawl.app cloud.",
|
|
6
6
|
"type": "module",
|
|
@@ -53,7 +53,7 @@
|
|
|
53
53
|
},
|
|
54
54
|
"dependencies": {
|
|
55
55
|
"@modelcontextprotocol/sdk": "^1.32.2",
|
|
56
|
-
"@webcrawl/client": "^1.
|
|
56
|
+
"@webcrawl/client": "^1.4.0",
|
|
57
57
|
"zod": "^4.0.0"
|
|
58
58
|
},
|
|
59
59
|
"engines": {
|