ONES Wiki MCP Server

Overview

The ONES Wiki MCP Server is a Spring‑based service that bridges the ONES collaboration platform with AI assistants through the Model Context Protocol (MCP). It solves a common pain point for developers: retrieving rich, structured content from an internal wiki and presenting it in a format that AI models can ingest effortlessly. By authenticating against the ONES instance, converting complex wiki URLs into simple API calls, and rendering HTML pages as clean Markdown, the server removes the friction of manual data extraction and formatting.

At its core, the service exposes a single MCP tool that accepts a full ONES wiki page URL and returns a well‑structured text representation. The conversion pipeline uses Jsoup to parse the page’s HTML, stripping out unsupported elements such as strikethroughs while preserving headings, lists, tables, images, and links. The resulting Markdown is deliberately AI‑friendly: it contains clear hierarchy, readable tables in key/value format, and descriptive alt text for images. This makes it straightforward for an assistant like Claude to parse the content, answer questions about it, or generate summaries without additional preprocessing.

Key capabilities include:

• Secure authentication via ONES credentials supplied in properties, command‑line flags, or environment variables, ensuring the server can access private wikis.
• Automatic URL translation, turning user‑friendly URLs into RESTful API endpoints without exposing internal logic to the client.
• Configurable output, allowing developers to tweak which elements are included or filtered, and to extend the service with additional tool methods such as page search or directory traversal.
• Seamless MCP integration, so the server can be added to any AI workflow that supports MCP, such as Claude Desktop or other Spring AI clients.

Typical use cases span from internal knowledge‑base querying—where a developer asks the assistant to fetch and explain a policy page—to automated documentation generation, where an AI tool pulls multiple wiki pages and composes a consolidated report. In continuous integration pipelines, the server can be invoked to validate that documentation stays up to date with code changes. Because it operates over HTTPS and manages session cookies internally, developers can integrate the service into existing security frameworks without exposing credentials.

What sets this MCP server apart is its end‑to‑end pipeline that abstracts away both authentication and HTML parsing. Developers no longer need to write custom scrapers or handle token management; the server presents a single, well‑documented tool that delivers ready‑to‑use Markdown. This focus on developer ergonomics, combined with the extensibility of Spring Boot and the standard MCP contract, makes the ONES Wiki MCP Server a valuable addition to any AI‑powered knowledge workflow.