July 10th, 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 is the second article in the series, and covers how you can install the Cline collaborative development assistant into Visual Studio Code (VS Code), connect it to Claude Sonnet v3.7 hosted in Amazon Bedrock, and add the new MCP Server for Spira.
Installing Cline into Visual Studio Code
Assuming that you already have Visual Studio Code (VS Code) installed, there are a variety of different software development agents available to you:
- Microsoft GitHub Copilot
- Amazon Q Developer
- Cline (open-source)
For our example, we are going to use Cline instead of Copilot or Amazon Q, however you could use Copilot or Amazon Q instead with the Spira MCP Server.
To install Cline, simply go to the Extensions section of VS Code and search for Cline:
Once you find it, click the option to install Cline into VS Code. Then restart VS Code and you should see the new Cline toolbar entry:
Next we have to connect Cline to our favorite LLM, in this case we will use Claude 3.7 Sonnet.
Connecting Cline to Claude Sonnet hosted in Amazon Bedrock
Since we are an Amazon Web Services (AWS) partner, we have access to all of the latest LLMs directly inside our Amazon Bedrock account. If you are not an AWS customer/partner, you could use Claude directly. Logging in to Amazon Bedrock, we have enabled model access to Claude Sonnet v3.7. This is the minimum version of Claude Sonnet that will work with Cline in agentic mode.
Once we have enabled model access to Claude Sonnet v3.7, we next need to create an Inference Profile and enable cross-region inference for Claude Sonnet 3.7. This will allow the maximum throughput of AI queries to be handled.
Make sure you copy the Inference Profile ID and Inference Profile ARN to the clipboard. You will need those pieces of information later.
Finally, you can go to the API Keys section and generate either a short-term or long-term API key:
You should check with your Chief Information Security Officer (CISO) about which type of API is appropriate for your use case.
You will need to record the API Key and matching Secret in a secure location such as your password manager.
Once you have Amazon Bedrock configured, go back to Cline and go to the Settings:
In this dialog box you need to choose your Api Provider. You can choose from a variety of options:
However for our example, we will choose Amazon Bedrock. Then we can enter in the Amazon Bedrock specific settings:
- AWS Access Key: Enter in the API Key you saved in your password manager
- AWS Secret Key: Enter in the API Secret that you saved in your password manager
- AWS Session Token: You can leave this blank
- AWS Region: Enter the AWS region you configured Bedrock access for earlier
- Use Cross Region Inference: Make sure this is checked
- Model: Enter in the full model name that is displayed in Bedrock
Once this is done, you can check that Cline can access the LLM API by trying out some sample tasks.
Assuming that is successful, the final step is to connect Cline to the Spira MCP Server
Connecting Cline to the Spira MCP Server
Firstly to add Spira as a new MCP server, you need to go to the MCP Server GitHub repository: https://github.com/Inflectra/mcp-server-spira
Then choose the option to Clone the repository to your local machine:
You can also just download the Zip file and unzip it to a location on your local computer.
Once you have cloned or unzipped the MCP server repository, you will need to install Python 3.11 or later and then run these commands:
# Clone the repository
git clone https://github.com/Inflectra/mcp-server-spira.git
cd mcp-server-spira
# Install in development mode
uv pip install -e ".[dev]"
# Install from PyPi
pip install mcp-server-spira
Now that you have installed the Spira MCP Server, go to Cline and click on the MCP Servers tab:
Click on the Edit Configuration button and you can now see the MCP Server configuration file:
You will need to edit the configuration file to add the Spira MCP Server:
{
"mcpServers": {
"inflectra-spira": {
"autoApprove": [
"get_my_incidents",
"get_products",
"get_test_cases"
],
"timeout": 60,
"command": "uv",
"args": [
"--directory",
"C:\\Git\\mcp-server-spira",
"run",
"main.py"
],
"env": {
"INFLECTRA_SPIRA_BASE_URL": "https://mycompany.spiraservice.net",
"INFLECTRA_SPIRA_USERNAME": "login",
"INFLECTRA_SPIRA_API_KEY": "{XXXXXX-XXXXXXX-XXXXXXX-XXXXXXX-XXXXX}"
},
"type": "stdio"
}
}
}
You will need to populate the following three special environment variables:
INFLECTRA_SPIRA_BASE_URL=The base URL for your instance of Spira (typically https://mycompany.spiraservice.net or https://demo-xx.spiraservice.net/mycompany)
INFLECTRA_SPIRA_USERNAME=The login name you use to access Spira
INFLECTRA_SPIRA_API_KEY=The API Key (RSS Token) you use to access the Spira REST API
Finally, to enable Spira MCP support in Cline, you simply need to restart VS Code and click on the Cline tab. You now should see the Spira MCP Server (inflectra-spira) listed under the list of installed and available MCP servers:
If you click on the chevron next to the MCP Server name, you can see a list of all the available MCP tools and resources:
You view the various tools and mark the ones that you are happy to run autonomously by selecting Auto-Approve = Yes. If a tool does not have that box checked, Cline will require you to explicitly approve each use of the tool.
Now that you have the Spira MCP Server installed, in the next article in the series, we'll discuss how you can use the MCP server to perform some common tasks.
