claude-code - 💡(How to fix) Fix Atlassian MCP: Add support for downloading Jira attachment content [1 comments, 2 participants]

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
GitHub stats
anthropics/claude-code#57845Fetched 2026-05-11 03:23:52
View on GitHub
Comments
1
Participants
2
Timeline
3
Reactions
1
Timeline (top)
labeled ×2commented ×1

Fix Action

Fix / Workaround

Current workaround: the assistant lists attachment filenames and Jira content URLs in a markdown file, and the user downloads them manually from the browser. This is functional but breaks the assistant's ability to read logs, analyse screenshots, or work with documents attached to tickets.

RAW_BUFFERClick to expand / collapse

Preflight Checklist

  • I have searched existing requests and this feature hasn't been requested yet
  • This is a single feature request (not multiple features)

Problem Statement

The Atlassian MCP connector can retrieve Jira issue data including attachment metadata (filename, size, content URL), but cannot download the attachment content itself. The fetch tool only accepts Atlassian Resource Identifiers (ARIs) — it does not accept the raw https://api.atlassian.com/.../attachment/content/... URLs returned in issue data. This affects both Claude Cowork (desktop) and Claude Code (CLI) users — anyone using the Atlassian MCP connector hits this gap regardless of how they connect.

Proposed Solution

Extend the Atlassian MCP connector to support downloading attachment content, either by:

Accepting attachment content URLs in the fetch tool (handling the Atlassian auth transparently), or Adding a dedicated get_attachment_content tool that takes an attachment ID or content URL and returns the file

Alternative Solutions

Current workaround: the assistant lists attachment filenames and Jira content URLs in a markdown file, and the user downloads them manually from the browser. This is functional but breaks the assistant's ability to read logs, analyse screenshots, or work with documents attached to tickets.

Priority

Medium - Would be very helpful

Feature Category

MCP server integration

Use Case Example

When building a project seed file from a Jira ticket (e.g. a bug with a stack trace screenshot or attached log file), the assistant can list the attachments but cannot read them. The user has to download each file manually, then re-upload or paste the contents. In a triage workflow where the key evidence is in an attachment, this requires an extra manual round-trip every time.

Additional Context

Related closed issue: #18359 — closed as a duplicate, but the duplicate chain appears to have no surviving open issue. If there is a canonical open issue, this could be a comment/upvote instead. Affects both Claude Cowork and Claude Code CLI users equally — the limitation is in the shared MCP connector, not surface-specific.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING