Troubleshooting
Super Productivity Not Running
Symptom: SP_UNAVAILABLE error from bridge commands.
Fix: Start the Super Productivity desktop app. The bridge requires SP to be running.
Local REST API Disabled
Symptom: Connection refused on port 3876.
Fix: In Super Productivity, go to Settings → Sync & Export → Local REST API and enable it. Restart SP if the setting was just changed.
MCP Host Cannot Find Command
Symptom: Host reports "command not found" or "spawn ENOENT" for sp-local-bridge-mcp.
Fix:
- Run
sp-local-bridge-print-config <host>— it outputs the absolute path - Verify the path exists:
ls -la /path/shown/sp-local-bridge-mcp - If missing, re-run
scripts/install.sh - Ensure your host config uses the absolute path, not a bare command name
GUI Host PATH Differs from Shell PATH
Symptom: Bridge works in terminal but host reports command not found.
Explanation: GUI-launched apps (VS Code, Claude Desktop) often don't inherit your shell PATH. The sp-local-bridge-print-config command outputs an absolute path specifically to avoid this issue.
Fix: Always use the absolute path from sp-local-bridge-print-config. If you used --bare, switch to the default (absolute) mode.
Tools Not Showing in Host
Symptom: Host is running but no Super Productivity tools appear.
Fix:
- Verify config file is saved in the correct location
- Restart the host application completely (not just reload)
- Check host logs for MCP server startup errors
- Run
sp-local-bridge-doctorto verify the bridge can start
Host Starts but Tool Calls Fail
Symptom: Tools appear but return errors when invoked.
Fix:
- Check if SP is running:
sp-local-bridge health - Check the specific error code:
SP_UNAVAILABLE→ SP not running or API disabledTIMEOUT→ SP is slow to respond, may be loadingINVALID_INPUT→ check the tool parameters you're passingTASK_NOT_FOUND→ the task ID doesn't exist
SP_UNAVAILABLE
Cause: The bridge cannot connect to http://127.0.0.1:3876.
Possible reasons:
- SP is not running
- Local REST API is disabled in SP settings
- SP is still starting up (wait a few seconds)
Recovery: Start SP and enable the API. The bridge will reconnect on the next request — no restart needed.
TIMEOUT
Cause: SP did not respond within the timeout window.
Possible reasons:
- SP is performing a heavy sync operation
- System is under high load
Recovery: Wait and retry. If persistent, restart SP.
Reinstall / Regenerate Host Config
If the bridge was reinstalled or moved:
# Regenerate config with new paths
sp-local-bridge-print-config <host>
# Update your host config file with the new snippet
# Restart the hostUninstall and Stale Config Cleanup
After uninstalling with scripts/uninstall.sh, remove any config snippets you added:
- VS Code Copilot: Remove the server entry from
.vscode/mcp.json - Claude Desktop: Remove the
super-productivityentry fromclaude_desktop_config.json - Codex: Remove the
[mcp_servers.superProductivity]section from~/.codex/config.toml
Stale config entries may cause the host to log errors on startup.