Traylinx Web Information Retrieval Agent MCP Endpoint

This Model Context Protocol (MCP) routing node serves as a proxy gateway to the operational Agentic Information Retrieval Service. It empowers client applications, such as Claude Desktop or Cursor, to leverage sophisticated search mechanisms, yielding both narrative digests and structured data representations (e.g., markup, imagery, etc.).

Available Functionality

query_web

Execute a comprehensive web information sweep utilizing the proprietary Traylinx API, which furnishes meticulous, highly contextualized results complete with verifiable sourcing.

Parameters: - information_request (string): The specific lexicon or query string directing the search operation. - temporal_constraint (string, optional): Designates a time window for result filtering. Valid options include: "month", "week", "day", or "hour". If omitted, results are returned without temporal restriction.

Operational Workflow

1. The local MCP instance must be provisioned with the Agentic Retrieval API endpoint URI and the requisite authentication token (typically via secure environment variable injection from the client configuration).
2. An MCP-aware consumer (e.g., Claude) dispatches a functional invocation to this server, specifying the search target and an optional time boundary.
3. This mediating server securely transmits the request, augmented with authorization credentials, to the upstream Agentic Retrieval Platform.
4. The system parses the heterogeneous payload received (which may contain narrative, HTML documents, aggregated search result entities, visual assets, or journalistic reports) and serializes it into a structured format consumable by the invoking client.

Deployment Prerequisites

System Requirements

• Node.js environment running version 18.0.0 or newer.
• A valid access credential (API Key) secured from Traylinx.com.

Phase 1: Credential Acquisition

1. Navigate to traylinx.com and establish an account.
2. Access the portal designated for developer resources or API management.
3. Generate a fresh API token specifically for the Agentic Information Retrieval Service.
4. Safeguard this key; it is mandatory for subsequent operational configuration.

Phase 2: MCP Node Setup

bash

Obtain the source code repository

git clone https://github.com/traylinx/traylinx-search-engine-mcp-server.git cd traylinx-search-engine-mcp-server

Install required dependencies

npm install

Compile the application artifact

npm run build

Phase 3: Client Integration Configuration

For Claude Desktop Users

Modify the configuration file located at claude_desktop_config.json:

{ "mcpServers": { "tlx-web-retrieval-agent-mcp-hub": { "command": "node", "args": ["path/to/traylinx-search-engine-mcp-server/dist/index.js"], "env": { "AGENTIC_SEARCH_API_KEY": "sk-lf-xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx", "AGENTIC_SEARCH_API_URL": "https://agentic-search-engines-n3n7u.ondigitalocean.app", "LOG_LEVEL": "INFO" } } } }

Access paths for the configuration file: - macOS: ~/Library/Application Support/Claude/claude_desktop_config.json - Windows: %APPDATA%\Claude\claude_desktop_config.json

For Cursor Users

Adjust the settings within your designated mcp.json file:

{ "tlx-web-retrieval-agent-mcp-hub": { "env": { "AGENTIC_SEARCH_API_KEY": "sk-lf-xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx", "AGENTIC_SEARCH_API_URL": "https://agentic-search-engines-n3n7u.ondigitalocean.app", "LOG_LEVEL": "INFO" }, "command": "node", "args": ["path/to/traylinx-search-engine-mcp-server/dist/index.js"] } }

CRITICAL NOTE: Substitute the placeholder authentication value with your actual secret obtained from Traylinx.com.

Validation Procedure

1. Following configuration adjustments, ensure the host MCP client application is subjected to a complete relaunch.
2. Initiate a novel conversational thread and direct the client to utilize the integrated mechanism:
3. "Invoke the web query function to ascertain facts regarding quantum mechanics."
4. "Search for the most recent intelligence on synthetic cognition, restricting returns to the past seven days."
5. "Retrieve and parse the textual content and underlying markup from the URI: https://traylinx.com"
6. When prompted for security confirmation, grant authorization.
7. You should subsequently observe a structured response encompassing both narrative substance and potentially machine-readable artifacts.

Advanced Operational Modalities

The Traylinx Retrieval Proxy supports the delivery of diverse data modalities:

• Textual Artifacts: Standardized Markdown formatting summarizing the retrieved insights.
• Embedded Markup: Source HTML content, particularly useful when fetching specific web pages.
• Search Entity Records: Structured objects detailing results, including title, canonical URL, and descriptive excerpts.
• Media Manifests: Pointers or descriptions for images, video streams, and analogous media discovered.
• News Feeds: Timely journalistic entries complete with associated graphical thumbnails and metadata.
• Raw Service Output: The complete, unmodified response payload from the backend API for highly granular inspection.

Utilizing Temporal Constraints

To confine results based on when they were published or indexed:

// Example invocation via Claude Desktop interface Access the web resource tool to locate contemporary reports on SpaceX operations, limiting results exclusively to those indexed within the last 24 hours.

// Example invocation from a custom client application { "name": "query_web", "arguments": { "information_request": "SpaceX orbital deployments", "temporal_constraint": "day" } }

Core Capabilities

• Multi-Format Output: Generates results spanning numerous content classifications beyond mere prose.
• Time-Bound Retrieval: Enables precise filtering of indexed material based on temporal proximity (hour, day, week, month).
• Secure Credential Management: Authentication tokens are strictly confined to the runtime environment variables.
• Endpoint Flexibility: The remote service address is configurable, allowing easy switching between staging and production backends.
• Full Protocol Adherence: Compliantly implements all mandatory specifications for an MCP server component.

Deployment via Smithery.ai Platform

This endpoint is readily containerizable and deployable on Smithery.ai:

1. Log in or create your Smithery platform account.
2. Select the option to 'Initiate Deployment of a New MCP Node'.
3. Input Identifier: tlx-web-retrieval-agent-mcp-hub
4. Specify Base Directory: . (indicating the root directory).
5. Click 'Provision'.

Once the service is live, it can be referenced within Claude's interface using the command:

Engage the tlx-web-retrieval-agent-mcp-hub to look up [your search subject]

Crucial Note: Ensure your security key (AGENTIC_SEARCH_API_KEY) is injected correctly within the Smithery environment configuration settings for this deployment.

Diagnostic Procedures

Should operational anomalies arise:

1. Verify that the API credential loaded into the configuration is error-free.
2. Confirm that the parent MCP client application has been fully recycled.
3. Check external network routing to confirm accessibility to the Agentic Retrieval Service.
4. For detailed introspection, elevate the LOG_LEVEL setting to DEBUG.

For direct technical assistance, contact the service originator at support@traylinx.com

Licensing

This repository is released under the terms of the MIT License (refer to the accompanying [LICENSE] file for particulars).