Skip to content

Troubleshooting

Common issues and solutions for NeuBird MCP.

Problem: mcp-server-neubird: command not found

Solutions:

  1. Check installation:

    Terminal window
    npm list -g mcp-server-neubird
  2. Reinstall globally:

    Terminal window
    npm install -g mcp-server-neubird
  3. Use npx instead:

    Terminal window
    npx mcp-server-neubird

Problem: Error: Node.js 20+ required

Solution:

Terminal window
# Check version
node --version
# Upgrade Node.js
# macOS
brew install node@20
# Linux
curl -fsSL https://deb.nodesource.com/setup_20.x | sudo -E bash -
sudo apt-get install -y nodejs

Problem: Authentication failed: 401 Unauthorized

Solutions:

  1. Verify credentials:

    Terminal window
    # Check environment variables
    echo $NEUBIRD_EMAIL
  2. Test login:

    • Try logging into NeuBird web UI
    • Reset password if needed
  3. Check for typos:

    • No extra spaces
    • No surrounding quotes
    • Correct email format
  4. For remote server (email/password):

    • Verify the X-NeuBird-Email and X-NeuBird-Password headers are correct (X-NeuBird-* headers are also accepted for backward compatibility)
    • Check the remote server URL is correct
  5. For remote server (bearer token):

    • Verify the token is not expired — tokens are JWTs with an exp claim
    • Test the token manually:
      Terminal window
      curl -s https://<your-deployment-name>.app.neubird.ai/mcp \
      -H "Authorization: Bearer <your-token>" \
      -H "Content-Type: application/json" \
      -H "Accept: application/json, text/event-stream" \
      -d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2025-03-26","capabilities":{},"clientInfo":{"name":"test","version":"1.0"}}}'
    • If expired, obtain a fresh token from the login endpoint (see Installation)

Problem: ECONNREFUSED or ETIMEDOUT

Solutions:

  1. Check network:

    Terminal window
    curl https://app.neubird.ai/api/health
  2. Check firewall:

    • Allow outbound HTTPS (443)
    • Check corporate proxy settings
  3. For remote server:

    • Verify the remote server URL is reachable
    • Check with your administrator

Problem: Can’t see NeuBird tools in Claude Desktop

Solutions:

  1. Verify config file location:

    Terminal window
    # macOS
    cat ~/Library/Application\ Support/Claude/claude_desktop_config.json
    # Windows
    type %APPDATA%\Claude\claude_desktop_config.json
  2. Check JSON syntax:

    • Use JSONLint
    • Look for missing commas
    • Check bracket matching
  3. Restart Claude Desktop:

    • Completely quit (Cmd+Q on macOS)
    • Wait 3 seconds
    • Relaunch
  4. Check logs:

    Terminal window
    # macOS
    tail -f ~/Library/Logs/Claude/mcp*.log
    # Look for errors

Problem: MCP server keeps crashing

Solutions:

  1. Test server directly:

    Terminal window
    npx @modelcontextprotocol/inspector npx mcp-server-neubird
  2. Check credentials:

    • Ensure all env vars are set
    • No special characters causing issues
  3. Update package:

    Terminal window
    npm update -g mcp-server-neubird

Problem: Investigation running for several minutes

Normal Duration: 30-90 seconds

If Longer:

  1. First investigation on new project:

    • Initial sync takes 5-10 minutes
    • Subsequent investigations faster
  2. Check connection sync:

    Show me connection status
  3. Complex incident:

    • Many data sources = longer time
    • This is normal

Problem: Investigation status returns ‘unknown’ or no progress data

Solutions:

  1. Investigation just started:

    • Wait 5-10 seconds for stream to initialize
    • Check again: Show me the investigation status
  2. Stream connection initializing:

    • Normal during first few seconds
    • Progress data will appear shortly
  3. Investigation hasn’t started yet:

    • Verify investigation was created successfully
    • Check session UUID is correct

Problem: investigation_summary.steps is empty or undefined

Solutions:

  1. Early initialization phase:

    • Investigation may be in early setup
    • Chain of thought messages haven’t arrived yet
    • Wait 30 seconds and check again
  2. Investigation just completed:

    • Use neubird_get_chain_of_thought for full step details
    • Use neubird_get_rca for final results
  3. Check investigation age:

    • Progress data is cleaned up after 1 hour
    • For older investigations, use other tools like get_chain_of_thought

Problem: Progress percentage hasn’t changed in several minutes

This is Normal When:

  1. Complex data queries:

    • Some steps require heavy database queries
    • Analysis phase often takes longer
    • May pause at 30-60% during correlation analysis
  2. Large data sets:

    • Processing many logs or metrics
    • Normal for thorough investigations
  3. Pattern detection:

    • ML-based analysis can take time
    • Expected behavior

Concerning If:

  1. Stuck at 0% for >2 minutes:

    • Investigation may have failed to start
    • Check connection sync status
    • Try creating a new investigation
  2. Stuck at same % for >5 minutes:

    • Check investigation status for errors
    • Look for error messages in status response
    • Contact support if issue persists

Problem: unique_sources only shows 1-2 sources, expected more

Solutions:

  1. Check connection sync:

    Show me project connections
    # Ensure all are in SYNCED state
  2. Verify data availability:

    Show me available resources for this project
    # Check if expected log/metric sources exist
  3. Time range issues:

    • Incident may be outside data retention period
    • Check if data exists for the incident timeframe
  4. Connection permissions:

    • Verify IAM/API key permissions
    • Ensure connections can access all resources

Problem: current_step field shows same message for several checks

Solutions:

  1. Check progress_percentage:

    • If increasing, investigation is progressing
    • current_step updates at major milestones only
    • Check completed_steps instead
  2. Long-running step:

    • Some steps naturally take longer
    • Analysis and diagnosis phases especially
    • Normal behavior
  3. Investigation may be done:

    • Check status field for “completed”
    • Get final RCA: Show me the RCA

Problem: RCA lacks details or specific recommendations

Solutions:

  1. Add context instructions:

    Create a SYSTEM instruction with our architecture details
  2. Add investigation steps:

    Create RCA instructions for common incident types
  3. Verify connections:

    Show me project connections
    # Ensure all are SYNCED
  4. Check data availability:

    Show me available resources for this project

Problem: RCA identifies wrong root cause

Solutions:

  1. Ask follow-up questions:

    Why do you think that's the root cause?
    What other possibilities exist?
  2. Add clarifying instruction:

    Create an instruction to guide this type of investigation better
  3. Test instruction first:

    • Apply to past session
    • Rerun investigation
    • Compare results

Problem: Connection stuck in SYNCING state

Solutions:

  1. Check status:

    Show me connection details for [connection-uuid]
  2. Verify credentials:

    • AWS: Check role ARN and external ID
    • Datadog: Check API/app keys
    • Azure: Check all 4 credentials
  3. Check permissions:

    • AWS: ReadOnlyAccess policy
    • Ensure trust relationship correct
  4. Wait longer:

    • First sync: 5-10 minutes normal
    • Large environments: up to 15 minutes

Problem: Expected resources not showing in discoveries

Solutions:

  1. Check sync status:

    Wait for connection to sync completely
  2. Verify region:

    • AWS: Ensure region included in connection
    • Resources in unconfigured regions won’t show
  3. Check permissions:

    • Missing permissions = missing resources
    • Verify IAM policies

Problem: Claude responses taking long time

Solutions:

  1. Use compact mode:

    Show me sessions in compact format
  2. Limit results:

    Show me only last 10 sessions
  3. Use global install:

    Terminal window
    npm install -g mcp-server-neubird
    # Faster than npx

Problem: MCP server using too much memory

Solutions:

  1. Restart MCP server:

    • Restart Claude Desktop
  2. Limit concurrent operations:

    • Investigate one alert at a time
  3. Update to latest version:

    Terminal window
    npm update -g mcp-server-neubird

Meaning: All alerts have been investigated OR filters are removing them

Solutions:

  1. Expand time range:

    Show uninvestigated alerts from last 30 days
  2. Check filters:

    Show me active FILTER instructions
  3. Include all severities:

    • Your filter might be too aggressive

Problem: Session UUID not found

Solutions:

  1. Verify UUID:

    • Check for typos
    • Use correct UUID format
  2. Check project:

    • Ensure using correct project
    • Session might be in different project
  3. List sessions:

    Show me all sessions

Problem: Connection not ready for use

Solution:

Wait for connection [uuid] to finish syncing

In Claude Desktop config:

{
"mcpServers": {
"neubird": {
"command": "npx",
"args": ["-y", "mcp-server-neubird"],
"env": {
"NEUBIRD_EMAIL": "...",
"NEUBIRD_PASSWORD": "...",
"NEUBIRD_LOG_LEVEL": "debug"
}
}
}
}

Test server in isolation:

Terminal window
npx @modelcontextprotocol/inspector npx mcp-server-neubird

Opens web interface showing:

  • Available tools
  • Tool execution
  • Request/response logs
  • Errors

If issues persist:

  1. Check logs:

    Terminal window
    # macOS
    tail -100 ~/Library/Logs/Claude/mcp*.log
  2. Test with inspector:

    Terminal window
    npx @modelcontextprotocol/inspector npx mcp-server-neubird
  3. Contact support:

  4. Documentation: