Background Setup
I’m working with Bitbucket Server connected to Jira through Application Links. We use a commit message validation hook to enforce proper formatting.
The Issue
Bitbucket automatically scans entire commit messages for Jira ticket references, not just the first line. This creates unwanted ticket associations when developers mention old tickets in commit details.
Example Scenario
Consider this commit message:
PROJ-5678: Update user authentication flow
This resolves the login timeout problem we discussed.
Similar approach was used in PROJ-9999 but needed adjustments
for the new security requirements.
Bitbucket links both PROJ-5678 and PROJ-9999 to this commit, even though PROJ-9999 is just a reference and not directly related to the actual changes.
What I Need
Is there a configuration option in either Bitbucket or Jira that restricts ticket detection to only the first line of commit messages? This would allow developers to mention related tickets in descriptions without creating false associations.
I’m using Bitbucket Server and Jira Server, but interested in solutions for Cloud versions too.
Nope, Bitbucket Server doesn’t have a setting to limit Jira ticket detection to just the first line. It scans the whole commit message by default. But I’ve run into this same issue and found a decent workaround.
We tweaked our commit message validation hook with a regex that only picks up ticket references in the first line. If someone puts a Jira ticket anywhere else in the message, the hook blocks the commit. Now devs have to be more careful about ticket references, but they can still mention related work by saying stuff like “similar to ticket PROJ-9999” or “based on work in PROJ-9999”.
Yeah, it makes commit messages a bit more restrictive, but it completely killed the false associations. You could also try using a custom Jira field to track related tickets instead of depending on commit message parsing.
Had this exact problem last year and found a workaround that actually works. Instead of fighting Bitbucket’s commit scanning, we just changed how we format our messages. Now devs put reference-only tickets in parentheses or write “mentioned in PROJ-9999” instead of bare ticket numbers. Bitbucket’s pattern matching is pretty predictable - it hunts for standard ticket formats, so you can break that pattern while keeping messages readable. We also moved related ticket info to our wiki instead of stuffing it all in commit messages. Takes some adjustment but completely killed the false linking without needing admin access or custom hooks.
there’s no easy fix for this in bitbucket or jira since it always checks the full commit msg. some teams just use prefixes like “ref:” for old tickets. that way, they stay out of the linking. another option is to turn off auto-linking, but that might be too much for most teams.