I work in the real world on real applications that are pretty large, complex and do more than just add a TODO item to a small list. Therefore I have been very skeptical about AI coming in and being this miracle that the AI companies would have you believe. (I have still seen no evidence of that btw. despite the recent mentions of AGI.)
Instead, I have seen AI as more of another tool to use in a developers’s toolkit. It can be a very useful tool in some circumstances but a tool none the less. Therefore, I like to make small incremental steps when introducing any AI into my workflow.
Recently I have been trying out the Atlassian MCP server. I have been quite impressed so far and think it can help in my existing workflow and make my daily developer life a little better.
The README is really well documented and it was easy to get it setup in Cursor. I got it connected to our Jira board that we use at work all within a few minutes.
Basically you have to:
- Generate an api token from Atlassian
- Pull the docker image that runs this MCP server
- Add the MCP tool to your Cursor config
{ "mcpServers": { "mcp-atlassian": { "command": "docker", "args": [ "run", "-i", "--rm", "-e", "CONFLUENCE_URL", "-e", "CONFLUENCE_USERNAME", "-e", "CONFLUENCE_API_TOKEN", "-e", "JIRA_URL", "-e", "JIRA_USERNAME", "-e", "JIRA_API_TOKEN", "ghcr.io/sooperset/mcp-atlassian:latest" ], "env": { "CONFLUENCE_URL": "https://your-company.atlassian.net/wiki", "CONFLUENCE_USERNAME": "your.email@company.com", "CONFLUENCE_API_TOKEN": "your_confluence_api_token", "JIRA_URL": "https://your-company.atlassian.net", "JIRA_USERNAME": "your.email@company.com", "JIRA_API_TOKEN": "your_jira_api_token" } } }}
The MCP server currently provides 42 tools which seems an awful lot and to be honest I haven’t tried them all out yet.
How I Use the Atlassian MCP to Scope Out Issues
There is a get_issue tool in this MCP and by pasting in the link to the issue, you can give your Cursor AI agent direct access to the whole contents of a specific issue in Jira.
This is really nice when there are a bunch of comments and descriptions. Saves you having to copy and paste a bunch of data.
I then ask the AI agent to review the ticket and make a Plan (only plan for now!), I have noticed the AI agent can be a bit keen sometimes to start cranking away at solutions before it really ‘understands’ the issue.
I then review the plan carefully and see if I agree with the plan and also if its even referencing the right area and files of the application. I like to take my time here because if the AI has the right context, the right files and the right instructions then it tends to do a pretty good job updating any code.
If you leave things too open or ambiguous then the AI agent has a tendency to start doing some random things which helps no one and wastes time and money.
Once it has come up with suggestions and I am happy with it I will either do the coding myself or let the AI have a go. It really depends on the issue and the context.
Creating Issues Through the Atlassian MCP
This is another great use case I have found. How often have you been working away in the code and find an issue that has nothing to do with the current thing you are working on. It’s too big an issue to fix in your current PR but it’s not big enough to be a major issue.
You make a mental note — “Make a ticket for that later” — because you don’t want to ruin your flow by leaving the editor and creating a Jira ticket, but then you never do and the bug silently persists in your code base.
With the Jira MCP in Cursor, you can just press cmd + i tag the file/s and ask the AI agent to create a Jira issue in the right project. This can be done quickly, you don’t lose your flow and the actual ticket content will likely be better than if you did it yourself as AI is good at writing these things.
Creating Proposals or RFCs in Confluence
We use Confluence in our company for certain documentation. Whilst this is great, the thought of writing out proposals or docs can sometimes be pretty uninspiring and tedious. It’s a very useful practice when working in a team but you don’t always have the motivation.
Again, this is where AI agents can shine. Being able to do this from your editor is really convenient and useful. It also makes it easier to reference existing code.
I use this to create initial drafts as I always find it easier to update and amend existing content rather than creating something from scratch.
Conclusion
I prefer the idea of incremental improvements in my workflows rather than big sweeping changes. I feel the Atlassian MCP is a great little addition to parts of my workflow that I previously found frustrating.
Creating accurate issues and documentation is an essential part of being a Software engineer and being able to have some useful help in this area makes this whole experience better.
Do you use the Atlassian MCP? If so, how do you use it? What other MCPs do you use and recommend?
Subscribe to My Weekly Updates!
Enjoyed This Post?
Enjoyed this one? Why then, hunt down that “clap” button and get clicking. It can be clicked up to 50 times, per reader!
If you found this blog post helpful, why not stay updated with my latest content? Subscribe to receive email notifications every time I publish.
What You’ll Get
By subscribing, you’ll get:
- Exciting Discoveries: Be the first to know about the latest tools and libraries.
- How-To Guides: Step-by-step articles to enhance your development skills.
- Opinion Pieces: Thought-provoking insights into the world of frontend development.
Join Our Community
I live in the vibrant city of Prague, Czech Republic, with my family. My blog is more than just articles; it’s a community of like-minded developers who share a love for innovation and learning.
About me
I’m a passionate Frontend Developer specialising in React and TypeScript. My professional journey revolves around exploring and mastering new tools and libraries within the JavaScript ecosystem.
Check out the original blog post on my blog.