Hey everyone,
I built a small open-source app called MCP Jira Automation. It uses MCP Atlassian to read Jira issues and helps automate API test workflows around them. The basic flow is: it reads a Jira issue, generates or updates API tests, runs them in Docker, opens a PR, and then comments the result back to Jira.
It supports GitHub, GitLab, Bitbucket, OpenAI, Anthropic, Gemini, vLLM, Aider, local Docker, and remote Docker over SSH. The SSH part is useful if you want to run the test automation on a separate dev/test server instead of your local machine.
I also added a sandbox mode, so the generated tests can be run in an isolated environment before anything is pushed or reported back. The goal is to make Jira → API tests → Docker run → PR a bit easier to manage.
Repo: https://github.com/ahmetguness/mcp-jira-automation
I’d love feedback from people using MCP, Jira, or API test automation. Does this workflow make sense? What would you improve?
Top comments (0)