feat: release 2.2.0 with PC-only tests and execution reliability fixes

Unify task scheduling and queue state, fix Android input and observation recovery, preserve screenshot paths, and match final review criteria accurately. Add collapsible copyable results and publish the complete 2.1.0 upgrade notes. Validation: 3189 backend tests, 326 frontend tests, 24 client tests, production build and quality gates.

artemis/agents/operator/prompts.py

@@ -586,6 +586,12 @@ def render_execution_incident(incident: dict, steps: list) -> str:
             " appeared on top of your target (dialog, sheet, banner, keyboard, or a"
             " re-laid-out screen)."
         )
+    elif category == "pixel_verification_unavailable":
+        lines.append(
+            "Evidence: target verification could not be completed; this does not show that"
+            " the target disappeared. Restore a working observation before locating the"
+            " target again. Do not infer a business application failure from this incident."
+        )
     elif category in ("target_disappeared", "pixel_target_disappeared"):
         prior_str = (
             f" Your last dispatched action was `{prior[0]}` (Step {prior[1]});"