Kiwi
Kiwi

Kiwi for Bitbucket

Have the agent on a Kiwi step open a Bitbucket Cloud pull request and read its pipeline results, through the Atlassian Rovo MCP Server.

  • Your agent works in Bitbucket through its own connection. Kiwi does not call Bitbucket and never holds its credentials.
  • A workflow skill can list Bitbucket among the tools it requires. When you supply your agent's available tools to a skill check and Bitbucket is not among them, the check reports SK15-UNSUPPORTED-TOOL.
  • A project Admin can add links to Bitbucket, with notes, under Connections and bind them to the steps that use them. The agent reads them with list_connections.
  • The agent can report the tools and plugins it ran with, including its Bitbucket connection, with report_configuration.
  • The agent submits links to its work in Bitbucket as the step's outputs. People review them, and sign-off stays with the people your workflow names.

What your agent can do in Bitbucket

  • List the repositories in a workspace, browse branches and read file contents.
  • Create a branch and commit changes to it.
  • Open a pull request, read its diff and comments, and comment on it.
  • Run a pipeline and read the log of a failed pipeline step.
  • Connect with a scoped API token, in a workspace linked to an Atlassian organization, once an organization admin enables API token authentication.

Set up

  1. Connect your agent to Kiwi

    Add Kiwi's hosted MCP server, https://kiwiflow.io/mcp, to your coding agent.

    Coding agentsQuickstart

  2. Connect your agent to Bitbucket

    Add the Bitbucket MCP server to your agent with this URL. The credentials stay with your agent.

    https://mcp.atlassian.com/v2/mcp

    Bitbucket documentation

  3. Example skill

    Publish instructions like these as a workflow skill that lists Bitbucket among its required tools.

    Take the Bitbucket workspace, repository and task from the assignment inputs.
Create a branch, commit the change and open a pull request.
Run the repository's pipeline and read the log of any failed step.
Submit the pull request link and the pipeline result as the step's outputs.
The reviewer named in the assignment approves and merges the pull request.
If the pipeline fails twice for reasons outside the change, stop and call
report_blocker with the log of the failed step.
Get startedRead the setup guide