oremus-web-search 0.1.3 → 0.1.4

package/README.md

@@ -90,6 +90,13 @@ Input:
 - `include_comments` (boolean, optional)
 - `include_tables` (boolean, optional)
 - `use_proxy` (boolean, optional)
+- `max_chars` (number, optional): cap returned text fields
+- `start_char` (number, optional): paging offset used with `max_chars`
+- `plain_text_fallback` (boolean, optional): if text/plain + extraction is empty, return raw body as text
+- `rewrite_github_blob_to_raw` (boolean, optional): rewrite GitHub `.../blob/...` URLs to `raw.githubusercontent.com/...`
+- `fetch_timeout_seconds` (number, optional): per-attempt HTTP timeout
+- `max_fetch_bytes` (number, optional): cap download size (may truncate HTML)
+- `max_total_seconds` (number, optional): best-effort overall time budget
 
 Output:
 

package/dist/index.js

@@ -7,7 +7,7 @@ import { StreamableHttpMcpClient } from "./streamableHttpMcpClient.js";
 import { webSearch } from "./searxng.js";
 import { FETCH_AND_EXTRACT_TOOL, ROTATE_VPN_TOOL, WEB_SEARCH_TOOL, isFetchAndExtractArgs, isRotateVpnArgs, isWebSearchArgs, } from "./types.js";
 import { createConfigResource, createHelpResource } from "./resources.js";
-const packageVersion = "0.1.3";
+const packageVersion = "0.1.4";
 async function main() {
     const config = loadConfig(process.env);
     const trafilaturaClient = new StreamableHttpMcpClient({
@@ -40,6 +40,15 @@ async function main() {
                 include_comments: args.include_comments ?? false,
                 include_tables: args.include_tables ?? false,
                 use_proxy: args.use_proxy ?? true,
+                max_chars: args.max_chars,
+                start_char: args.start_char,
+                max_fetch_bytes: args.max_fetch_bytes,
+                fetch_timeout_seconds: args.fetch_timeout_seconds,
+                user_agent: args.user_agent,
+                accept_language: args.accept_language,
+                plain_text_fallback: args.plain_text_fallback,
+                rewrite_github_blob_to_raw: args.rewrite_github_blob_to_raw,
+                max_total_seconds: args.max_total_seconds,
             });
             return { content: result.content };
         }

package/dist/streamableHttpMcpClient.js

@@ -112,7 +112,7 @@ export class StreamableHttpMcpClient {
                 params: {
                     protocolVersion: "2025-11-25",
                     capabilities: {},
-                    clientInfo: { name: "oremus-web-search", version: "0.1.3" },
+                    clientInfo: { name: "oremus-web-search", version: "0.1.4" },
                 },
             };
             const initResp = await this.post(initReq);

package/dist/types.js

@@ -67,6 +67,42 @@ export const FETCH_AND_EXTRACT_TOOL = {
                 description: "Whether the Trafilatura service should route via its proxy pool.",
                 default: true,
             },
+            max_chars: {
+                type: "number",
+                description: "Maximum number of characters to return for extracted text fields.",
+            },
+            start_char: {
+                type: "number",
+                description: "Starting character offset for extracted text fields (used with max_chars).",
+            },
+            max_fetch_bytes: {
+                type: "number",
+                description: "Maximum number of bytes to download before extraction (may truncate HTML).",
+            },
+            fetch_timeout_seconds: {
+                type: "number",
+                description: "HTTP fetch timeout in seconds (per attempt).",
+            },
+            user_agent: {
+                type: "string",
+                description: "Override User-Agent header for the upstream fetch.",
+            },
+            accept_language: {
+                type: "string",
+                description: "Optional Accept-Language header for the upstream fetch.",
+            },
+            plain_text_fallback: {
+                type: "boolean",
+                description: "If the upstream is text/plain and extraction is empty, return the raw body as text.",
+            },
+            rewrite_github_blob_to_raw: {
+                type: "boolean",
+                description: "If the URL is a GitHub blob page, rewrite to raw.githubusercontent.com before fetching.",
+            },
+            max_total_seconds: {
+                type: "number",
+                description: "Maximum total time budget in seconds for all attempts (best-effort).",
+            },
         },
         required: ["url"],
     },

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "oremus-web-search",
-  "version": "0.1.3",
+  "version": "0.1.4",
   "description": "MCP server that combines SearXNG web search with Trafilatura extraction",
   "license": "MIT",
   "type": "module",