Atlassian MCP Server Rate Limits

[Craig Dawson]

When I use the Atlassian MCP Server to query tickets, it very quickly hits a rate limit. I did a query to list cards in an Epic (there were 30 cards in the result set) and then it wouldn't let me do any more queries for at least 30 minutes. I was then able to update the description on one card, and then hit the rate limit again so couldn't do anything else.

I was using the MCP server OK last week and wasn't getting this issue, so what is happening?

The MCP server response also doesn't contain information about how long to wait for the rate limit to reset, or how many requests I have used up.

[Daniel Schuba]

I also observed this yesterday evening for the first time. It also looks like the header information from the request about when to retry are not passed to give the LLMs a hint on how long to wait so they are just guessing.

This currently blocks my work a lot and I haven't though that some searches, reading 5 stories and adding comments to them would trigger rate limits so fast.

[aaron_jorgensen]

This is also significantly affect our company.