[BUG] Atlassian MCP updateConfluencePage hangs and fails with large content payloads
Resolved 💬 7 comments Opened Jun 10, 2025 by dkoopman-godaddy Closed Jan 3, 2026
Environment
- Platform (select one):
- [ ] Anthropic API
- [x] AWS Bedrock
- [ ] Google Vertex AI
- [ ] Other:
- Claude CLI version: 1.0.18
- Operating System: macOS (Darwin 23.6.0)
- Terminal: iTerm2 3.5.14
Bug Description
The mcp__atlassian__updateConfluencePage tool hangs indefinitely and eventually fails with "API Error 1/10 ... 10/10" when attempting to update a Confluence page with large content payloads (~58,000 characters). The same operation works successfully with smaller content (first 33 lines, 3,268 characters).
Steps to Reproduce
- Set up Atlassian MCP server integration with Claude Code (https://mcp.atlassian.com/v1/sse)
- Attempt to use
mcp__atlassian__updateConfluencePagewith a large markdown content body (approximately 58,000 characters, 74 lines) - Observe the tool hanging for extended periods
- Eventually receive "API Error 1/10 ... 10/10" failure messages
- Retry the same operation with only the first 33 lines of content (3,268 characters) - this works successfully
Expected Behavior
- The tool should either successfully update the Confluence page regardless of content size (within reasonable limits)
- OR provide a clear error message indicating content size limitations
- OR fail quickly with a descriptive error rather than hanging
Actual Behavior
- Tool hangs for extended periods when processing large content
- Eventually fails with generic "API Error 1/10 ... 10/10" messages
- No clear indication of what the size limitation is or that size is the issue
- Smaller content works fine, suggesting an undocumented size threshold somewhere between 3,268 and 58,000 characters
Additional Context
- Issue appears to be related to content size rather than content type (markdown formatting)
- Current limitation: Only workaround is to use truncated content, which is inadequate when full page updates are required
- This suggests there may be an internal buffer limitation or timeout in the MCP integration
- The hanging behavior makes it difficult to quickly identify the root cause
- No clear documentation found regarding content size limits for this tool
- MCP server reference: https://mcp.atlassian.com/v1/sse
This issue has 7 comments on GitHub. Read the full discussion on GitHub ↗