chore: remove VibeCode_TWS staging files

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Author: Jerrycyborg

.ai/handoff/NEXT_ACTIONS.md

@@ -6,15 +6,16 @@ Priority order: top to bottom. Each task is self-contained -- start without aski
 
 ## Phase 3: Production Readiness & Scaling (v0.2-alpha → v0.3)
 
-| ID    | Task                                                    | Priority | Est   |
-| ----- | ------------------------------------------------------- | -------- | ----- |
-| T-004 | Implement Redis atomic task claiming (Lua script)       | HIGH     | 1 day |
-| T-005 | Add soak test framework (2-3 nodes, 100+ tasks)         | HIGH     | 1 day |
-| T-006 | Multi-instance control-plane load balancing tests       | MEDIUM   | 1 day |
-| T-007 | Add Grafana dashboard templates for /metrics            | MEDIUM   | 1 day |
-| T-008 | Kubernetes deployment manifests + Helm chart            | MEDIUM   | 2 day |
-| T-009 | Add comprehensive E2E test suite                        | MEDIUM   | 1 day |
-| T-010 | Performance benchmarks (task throughput, claim latency) | LOW      | 1 day |
+| ID    | Task                                                    | Priority | Status      | Notes                                          |
+| ----- | ------------------------------------------------------- | -------- | ----------- | ---------------------------------------------- |
+| T-004 | Investigate test flakiness                              | HIGH     | NOT_STARTED | Loop tests fail, single runs pass              |
+| T-005 | Complete Redis atomic task claiming (Redlock)           | HIGH     | IN_PROGRESS | Basic impl works; needs Redlock for multi-inst |
+| T-006 | Add soak test framework (2-3 nodes, 100+ tasks)         | HIGH     | NOT_STARTED |                                                |
+| T-007 | Multi-instance control-plane load balancing tests       | MEDIUM   | NOT_STARTED | Depends on T-005 completion                    |
+| T-008 | Add Grafana dashboard templates for /metrics            | MEDIUM   | NOT_STARTED |                                                |
+| T-009 | Kubernetes deployment manifests + Helm chart            | MEDIUM   | NOT_STARTED |                                                |
+| T-010 | Add comprehensive E2E test suite                        | MEDIUM   | NOT_STARTED |                                                |
+| T-011 | Performance benchmarks (task throughput, claim latency) | LOW      | NOT_STARTED |                                                |
 
 ---
 

.ai/handoff/STATUS.md

@@ -1,7 +1,8 @@
 <!-- SECTION: summary -->
 
 OpenClaw EdgeMesh control plane: all 6 milestones + T-001/002/003 + production hardening shipped.
-89/89 tests pass, build is clean. Ready for v0.2-alpha stabilization.
+89/89 tests pass in single runs. Test flakiness detected in loop runs (T-004 to investigate).
+Ready for v0.2-alpha stabilization, moving to Phase 3 production hardening.
 
 <!-- /SECTION: summary -->
 
@@ -80,7 +81,8 @@ OpenClaw EdgeMesh control plane: all 6 milestones + T-001/002/003 + production h
 
 | Gap                                    | Severity | Task                |
 | -------------------------------------- | -------- | ------------------- |
+| Test flakiness (loop vs single runs)   | medium   | T-004 (investigate) |
 | SSE fan-out only within single process | low      | future              |
-| Redis claimTask race condition         | medium   | Issue #2 (deferred) |
+| Redis multi-instance atomicity         | medium   | T-005 (Redlock)     |
 
 <!-- /SECTION: gaps -->