Content
View differences
Updated by Niels Lindenthal 3 months ago
**As a:** # User Problem
## User
* Project managers who want to query, summarize, or report on projects naturally using AI tools.
* Product owners who want to automate repetitive project reporting.
* Project members who need quick summaries, comparisons, or contextual insights from project data.
* Admins who want to safely connect OpenProject data with AI assistants.
<br>
##
Problem
* Users want to interact with OpenProject using natural language with external AI assistants to generate project status reports, find work packages, or analyze progress.
* Currently, they must extract data manually, build API scripts, or rely on nonstandard integrations.
* There’s no consistent, secure, standard protocol to let LLMs work with OpenProject data.
<br>
<br>
##
**I want:** Pain
* Manually copy project data for analysis
* Use API tokens in custom scripts which is a technical barrier and not maintainable.
* AI models can’t access project data directly
<br>
# Business Case
## Solution
* Implement a Model Context Protocol (MCP) server within OpenProject, using the official Ruby MCP SDK.
* Use HTTP transport only (no WebSocket or streaming) for simpler infrastructure and easier scaling.
* Reuse OpenProject’s existing OAuth2 authentication (with PKCE) for secure access — compatible with both browser-based and desktop MCP clients.
* Provide two search tools out of the box:
* searchProject: filters projects by standard attributes (name, identifier, status).
* searchWorkPackage: filters work packages by standard attributes (subject, type, status, project, version, user).
* Expose resources derived from OpenProject APIv3:
* projects
* work\_packages
* versions
* users
* Add an MCP admin page under Administration -&gt; AI Integration:
* Allows renaming or localizing tool/resource names and descriptions.
* Enables or disables specific tools/resources for compliance or privacy.
<br>
## Differentiation
* One of the first project management platform with native MCP support, enabling direct AI assistant access to real project data.
* It to be fully Fully integrated with OpenProject’s existing authentication and permission system. permissions.
* An admin Admin customization layer to align AI vocabulary with organizational terminology (e.g., “features,” “tickets,” or “projectOne”).
* A simple and easy to operate Simpler architecture (HTTP-only, no streaming) and easy easier deployment in my own infrastructure. for on-premise customers.
<br>
##
**Messaging:** Next iteration
* Add prompt support for standardized templates (e.g., “Project Status Report,” “Sprint Summary,” “Risk Overview”).
* Extend the MCP admin UI to manage prompt templates and output schemas.
* Add semantic search for projects and work packages.
* Optional support for streaming / real-time notifications.
* Expand resource catalog (meetings, comments, attachments).
<br>
# Launch and Growth
## Measures
* Number of successful MCP client connections.
* Time saved on automated workflows vs. manual workflows.
* Positive feedback from pilot customers integrating with AI agents.
<br>
## Messaging
<figure class="table op-uc-figure_align-center op-uc-figure"><table class="op-uc-table"><tbody><tr class="op-uc-table--row"><th class="op-uc-table--cell op-uc-table--cell_head"><p class="op-uc-p">Headline</p></th><td class="op-uc-table--cell"><p class="op-uc-p">Mercedes-AMG, Deutsche Bahn, and OpenProject have joined forces to develop AI technology that class="op-uc-p">OpenProject connects the open source project management system OpenProject directly with AI assistants through the Model Context Protocol (MCP).</p></td></tr><tr Protocol</p></td></tr><tr class="op-uc-table--row"><th class="op-uc-table--cell op-uc-table--cell_head"><p class="op-uc-p">First Paragraph</p></th><td class="op-uc-table--cell"><p class="op-uc-p">With class="op-uc-p">OpenProject now integrates with AI tools via the MCP Server, open Model Context Protocol (MCP). This allows users can to securely query, summarize, and report on real project data by using natural language directly in their AI agent.</p></td></tr><tr language, without custom scripts or manual exports.</p></td></tr><tr class="op-uc-table--row"><th class="op-uc-table--cell op-uc-table--cell_head"><p class="op-uc-p">Customer Quote</p></th><td class="op-uc-table--cell"><p class="op-uc-p">-</p></td></tr></tbody></table></figure><br> class="op-uc-p">-</p></td></tr></tbody></table></figure>
* Product owners who want to automate repetitive project reporting.
* Project members who need quick summaries, comparisons, or contextual insights from project data.
* Admins who want to safely connect OpenProject data with AI assistants.
<br>
##
* Users want to interact with OpenProject using natural language with external AI assistants to generate project status reports, find work packages, or analyze progress.
* Currently, they must extract data manually, build API scripts, or rely on nonstandard integrations.
* There’s no consistent, secure, standard protocol to let LLMs work with OpenProject data.
<br>
<br>
**I want:**
*
* Use API tokens in custom scripts which is a technical barrier and not maintainable.
* AI models can’t access project data directly
<br>
# Business Case
## Solution
* Implement a Model Context Protocol (MCP) server within OpenProject, using the official Ruby MCP SDK.
* Use HTTP transport only (no WebSocket or streaming) for simpler infrastructure and easier scaling.
* Reuse OpenProject’s existing OAuth2 authentication (with PKCE) for secure access — compatible with both browser-based and desktop MCP clients.
* Provide two search tools out of the box:
* searchProject: filters projects by standard attributes (name, identifier, status).
* searchWorkPackage: filters work packages by standard attributes (subject, type, status, project, version, user).
* Expose resources derived from OpenProject APIv3:
* projects
* work\_packages
* versions
* users
* Add an MCP admin page under Administration -&gt; AI Integration:
* Allows renaming or localizing tool/resource names and descriptions.
* Enables or disables specific tools/resources for compliance or privacy.
<br>
## Differentiation
*
* It to be fully
* An admin
* A simple and easy to operate
<br>
##
**Messaging:**
* Extend the MCP admin UI to manage prompt templates and output schemas.
* Add semantic search for projects and work packages.
* Optional support for streaming / real-time notifications.
* Expand resource catalog (meetings, comments, attachments).
<br>
# Launch and Growth
## Measures
* Number of successful MCP client connections.
* Time saved on automated workflows vs. manual workflows.
* Positive feedback from pilot customers integrating with AI agents.
<br>
## Messaging