| name | pivot-to-successful-diagnostic-as-new-baseline |
| description | Use when a diagnostic method succeeds where the primary method failed. Switch to the successful method as the new working baseline. |
| category | robotics/execution |
Pivot to Successful Diagnostic as New Baseline
- When a diagnostic succeeds (e.g., direct WebSocket connects), note this as the working baseline.
- The failure is now isolated to whatever the successful method bypasses (e.g., sidecar, proxy, auth layer).
- Stop retrying the failing method. Instead, investigate the difference between the successful and failing approaches.
- Report to the user: "Direct connection works, so the robot is reachable. The issue is specific to [sidecar/proxy]."
Example:
- Sidecar + WebSocket → fails, falls back to mock.
- Direct Python WebSocket → succeeds.
- Pivot: Debug sidecar WebSocket client code, not robot connectivity.
Anti-pattern: Saying "let's try sidecar again" after discovering direct connection works, without investigating why the discrepancy exists.