July 9th, 2025 by Adam Sandman
In this blog series we are introducing our new MCP Server for Spira, and demonstrating how you can use it to turbocharge your development processes. Model Context Protocol (MCP) is a new technology for connecting AI tools with backend services. This blog accompanies the video where I demonstrate the integration between our Spira project management platform, and Visual Studio Code (VS Code) running the Cline AI assistant. We showcase the system's ability to retrieve, analyze, and interact with project data using natural language queries, highlighting features such as task management, requirement analysis, and test case handling. As you will see in both the blog and live video, the AI-powered tool has an almost uncanny capability to intelligently interpret user intent when using the MCP server. This intelligence lets it combine multiple data sources, and perform complex operations autonomously, offering a significant improvement over traditional IDE integrations and plugins.
This first part of the blog series includes the Zoom recording and a summary of how you can use Spira with VS Code and Cline. The later parts will cover the setup in more detail as well as showcasing some of the use cases.
What is MCP and Why Create a Server for Spira?
As described in our background paper - Model Context Protocol (MCP) is a new AI "middleware" technology that lets you connect different AI tools and agents with a variety of with backend services:
In the context of the software development lifecycle, these different backend services will usually be a combination of SaaS products (e.g. Spira, GitHub, Slack, AWS Console, etc.) that are accessible through an API as well as local data storage such as Git repositories:
When developers are working in their Integrated Development Environment (IDE) such as Visual Studio Code, Cursor or Windsurf, it is convenient for the developers to be able to use their AI assistant to not only help write the code for the application they are building (or even build the entire application in the case of Vibe Coding), but to also interact with external services such as Spira. For example, they might want to ask their assistant to get a list of all the tasks they are working on, or get a list of all the bugs they have to fix:
What Features and Tools Does the Spira MCP Server Provide?
The Inflectra MCP Server for Spira includes support for a variety of common tasks, including:
My Work
This feature provides easy access to the list of artifacts that have been assigned to the current user- My Tasks: Provides operations for working with the Spira tasks I have been assigned
- My Requirements: Provides operations for working with the Spira requirements I have been assigned
- My Incidents: Provides operations for working with the Spira incidents I have been assigned
- My Test Cases: Provides operations for working with the Spira test cases I have been assigned
- My Test Sets: Provides operations for working with the Spira test sets I have been assigned
Workspaces
This feature provides tools that let you retrieve and modify the different workspaces inside Spira- Programs: Provides operations for working with Spira programs
- Products: Provides operations for working with Spira products
- Product Templates: Provides operations for working with Spira product templates
Program Artifacts
This feature provides tools that let you retrieve and modify the different artifacts inside a Spira program- Capabilities: Provides operations for working with the Spira capabilities in a program backlog
- Milestones: Provides operations for working with the Spira milestones in a program
Product Artifacts
This feature provides tools that let you retrieve and modify the different artifacts inside a Spira product- Requirements: Provides operations for working with the Spira requirements in a product
- Releases: Provides operations for working with the Spira releases in a product
- Test Cases: Provides operations for working with the Spira test case folders and test cases in a product
- Test Sets: Provides operations for working with the Spira test set folders and test sets in a product
- Test Runs: Provides operations for working with the Spira test runs in a product
- Tasks: Provides operations for working with the Spira tasks in a product
- Incidents: Provides operations for working with the Spira incidents (e.g. bugs, enhancements, issues, etc.) in a product
- Automation Hosts: Provides operations for working with the Spira automation hosts in a product
Template Configuration
This feature provides tools that let you view and modify the configuration and settings of Spira product templates- Artifact Types: Retrieves information on the artifact types in a product template, and their sub-types
- Custom Properties: Retrieves information on the artifact types in a product template, and their custom properties
Automation
This feature provides tools that let you integrate automated DevOps tools such as test automation frameworks and CI/CD pipelines- Automated Test Runs: Provides operations for recording automated test run results into Spira
- Builds: Provides operations for recording the results of CI/CD builds into Spira
The good news is that the MCP Server for Spira is fully open-source (MIT License) so as customers find additional needs that it doesn't currently support, they can suggest the additional feature or even built it themselves and submit a pull request. In addition, we have provided a Claude.MD file that will help Claude (and other agentic LLMs) create new tools and features automatically! We will discuss this aspect in more detail in a subsequent article.
Which Use Cases Are Covered in the Video?
In the video that accompanies this article, we demonstrate first how to use the Cline AI assistant integrated with VS Code to retrieve and analyze assigned tasks and incidents using natural language queries. The AI assistant can fetch live data from Spira, perform initial analysis, and provide insights such as critical and overdue tasks. It also explains the auto-approve feature for repeated tasks and demonstrates how the AI can combine information from multiple queries to provide comprehensive summaries. We highlight the contextual nature of the interactions and the ability to drill down into specific details, such as identifying critical bugs or potential risks in unclassified items.
After this initial use case, we explore some of the additional capabilities of the MCP server, focusing on workspaces and programs in Spira. We demonstrate how to retrieve a list of products and programs that the user has access to The system intelligently groups and summarizes the products and programs.
Next, we show how the system can group a list of programs by industry and retrieve requirements for a specific product. The system successfully retrieves and categorizes requirements for the product, grouping them by type and providing metadata about priorities and statuses. We then show how you can request to filter the requirements to show only those of high and critical importance without needing to make a separate API call.
To show the versatility of the MCP Server when coupled with a powerful agentic tool such as Cline, we demonstrate fetch requirements and test cases from different projects, even when given slightly imprecise project names. The agent can intelligently combine multiple MCP tools, such as getting product lists and requirements, and can work autonomously with auto-approve enabled.
Finally, we showcase how the agent can retrieve and analyze test case data, record test results, and perform multi-stage operations. We highlights the system's ability to autonomously execute multiple API calls and combine data from different sources.
Summary
The integration of Developer Tools such VS Code, Cursor, etc. with AI assistants like Copilot, Cline and AmazonQ, using the power of a Large Action Model (LAM) such as Anthropic Claude Sonnet allows users to ask questions in natural language from the "comfort" of their IDE, get real-time data from Spira (and other systems), then interpret that data before presenting it to the user so that it is meaningful and actionable. In addition to retrieving data, the agents can suggest actions that the user can take (such as updating statuses, recording test results) and if auto-approve is enabled, even take action autonomously.
In the next article in this series, we shall show you how you can setup a simple Agentic environment using Visual Studio Code (VS Code), Cline and Claude Sonnet running on Amazon Bedrock.