Updates documentation for node-banana PR #42.

Changes Made

Changelog (pages/changelog.mdx)

• Added entry under "Fixed" section for version 1.3.0
• Documents fix for race condition causing duplicate workflow executions when rapidly triggering execution

Summary of PR #42

PR #42 fixed a race condition in workflow execution by moving the isRunning guard before async logger calls in both executeWorkflow() and regenerateNode() methods. This prevents concurrent calls from passing the guard during the async gap, which could cause duplicate workflow executions when users rapidly click the execute button.

The fix includes 4 new integration tests verifying:

1. Concurrent executeWorkflow() calls only execute nodes once
2. isRunning is set synchronously before any await
3. Concurrent regenerateNode() calls only execute once
4. Multiple disconnected nodes produce exactly the expected number of fetch calls

🤖 Generated with Claude Code