What is MCP, How Can It Be Used, & How Is Inflectra Building It?
MCP might be the next frontier when it comes to AI in software development. But what does this new protocol actually do, how can you use it in your development processes, and how is Inflectra already starting to build for it? Keep reading to learn more.
What is MCP (Model Context Protocol)?
MCP, or Model Context Protocol, is an open, JSON-RPC 2.0–based standard that Anthropic (the creators of Claude.ai) released in November 2024. Since then, it’s quickly gained traction with industry leaders, being adopted by OpenAI, Google, Microsoft, Amazon AWS, and more. The goal of MCP is to facilitate frictionless interoperability so developers can build an integration once, and any compliant model or agent in the future can use it without additional glue code.
In essence, think of MCP as a “USB-C for AI” — it gives LLMs a uniform, permissioned way to read structured context, call functions, and push results back to any systems with an MCP server. This interconnectivity is incredibly valuable for modern software and AI companies alike, positioned to become the connective tissue for next-gen agentic workflows.
How Does MCP Work?
An MCP server (e.g. for GitHub, Slack, Spira, etc.) starts by registering resources and JSON-schema-defined methods. An MCP client (such as an LLM agent) then queries the MCP server to discover capabilities and can invoke actions via JSON-RPC.
The MCP server enforces authentication, executes the call (e.g. read data, write code, run test, etc.), and returns a typed result. This is purposely designed to mirror REST so existing API gateways, observability stacks, and security scanners can easily slot in.
How Can it be Used in Software Development?
MCP might be able to speed up and enhance your development pipeline in a number of ways. From pulling requirement and defect information in order to generate missing test cases, to using an agent to read failed tests, refactor the code, and re-run tests, there are a multitude of creative ways that developers are integrating MCP.
From DevOps teams to low-code SaaS vendors, here are some examples of how early adopters are using MCP:
- Some coding assistants are starting to be able to call MCP servers for Git, issue trackers, or log storage when a user asks, “Why did this test fail?” to propose and push a fix (all from within the editor).
- Cutting-edge software testing tools are exposing their test suites to “Test Runner” MCP servers so agents can fetch current coverage of metrics via the MCP, then highlight missing assertions or newly added code paths for regression testing.
- CI scaffolding can be improved if an agent detects the pipeline configuration via a Git MCP server, automatically scaffolding CI steps (e.g. run lint, run tests, notify) and pushing the updated CI file as a PR so non-expert users can adopt CI workflows.
Software developers of existing apps are also leveraging MCP to add features like:
- Cross-app automation in collaboration tools that use MCP servers, where an agent can translate, “Invite everyone who bought tickets this week to Slack” into calls, read Shopify orders, build messages, and post them.
- Organizational apps like calendars are starting to incorporate MCP servers to perform actions like rescheduling tomorrow’s meeting or adding someone to an invite without the user having to open yet another tab or program.
- Performance monitoring platforms are adding agents that can connect with CRM MCP servers to pull purchase data, draft messages based on the data, connect to Slack or Gmail MCPs, and send these alerts to relevant users.
How is Inflectra Integrating MCP?
At Inflectra, we’re working to keep our partners ahead of the competition and on the cutting edge of new technologies.
One of our customers had already created a community MCP Server for Spira using a combination of TypeScript and JavaScript, so we realized that the time was ripe for an official Inflectra MCP Server.
We’ve released an official MCP server for Spira using Python, and there are a few ways we’re already building (or considering building) this innovation into our products to enhance your workflows:
- The My Work feature lets developers request their currently assigned artifacts and work on them from the comfort of their favorite IDE (e.g. Visual Studio Code)
- The Automation feature lets DevOps engineers push test automation results and CI/CD builds to Spira using the MCP Server automation tools
- The Product Artifacts feature lets managers access all of the artifacts in a product (requirements, test cases, test sets, tasks, incidents, releases and risks) and perform analytics and reporting using natural language queries.
- The Program Artifacts feature lets managers access all of the artifacts in a program (capabilities and milestones) and perform analytics and reporting using natural language queries.
- The Workspaces feature provides tools that let you retrieve and modify the different workspaces inside Spira, including products, programs, and product templates
- The Template Configuration feature provides tools that let you view and modify the configuration and settings of Spira product templates
Personal Helpers
When connected to the Spira MCP, AI assistants can pull information for anything that’s assigned to you without clicking through menus or typing IDs. You could ask, “Show me everything I’m responsible for,” and your agent will fetch all artifacts assigned to you so you can see a single, prioritized to-do list.
You could also ask your AI assistant to “List my open bugs,” or “Show me my outstanding requirements,” and our MCP enables your AI to read, update statuses, or add comments without leaving the interface.
Project Management
Looking at the bigger picture, our Spira MCP server also allows you to check on progress, releases, and how everything is working together. You could say, “Give me the list of all environments or teams,” or “Show me this program’s milestones,” and your assistant is able to help you gather and coordinate around this information. We’re still adding more and more functionality, but you can already add new custom fields to a template, just by asking.
Automated Test Execution
One of the additional functionalities we’re looking to add to our MCP servers is to have Rapise run automated tests on specified components, simply by asking your preferred AI assistant to do so. This will enable developers and testers to be more efficient, not only in reducing context switching between systems but also in designating what pieces of code they want to test using natural language.
Upgrade Your Development Infrastructure Today
We’re constantly working to improve our development platforms so you always have the most advanced tools at your fingertips. As vibe coding is gaining popularity in development circles, it is ushering in a new era that is powered by AI and AI tools.
Now is the time to invest in tools to future-proof your pipelines and QA processes with platforms like Spira and Rapise. Sign up for a free 30-day trial to get hands-on with our industry-leading tools and see why partners across the globe trust our systems to deliver quality software on time.
