CYA-WP-0005 T05 done (ralph iter 5): Minimal Profile 1 (Reflexion verbal) spike

- Added KIND_REFLECTION + remember_reflection() helper in memory (thin, exported).
- Wired into run_retrospection(): optional 'capture verbal lesson' step at end.
- Main recall now includes reflections for preferential activation.
- Final LLM response + explain surface verbal reflections when activated.
- Added roundtrip test + import updates.
- Small README note.
- All changes small/inspectable, safety preserved (still through RiskClassifier).
- T05 acceptance criteria met with working end-to-end spike.

Committed as ralph iter 5. Ready for T06+ or close.

workplans/CYA-WP-0005-agentic-memory-profiles-and-phase-memory-feedback.md

@@ -137,9 +137,11 @@ completed: "2026-05-28 ralph iter 4"
 
 ```task
 id: CYA-WP-0005-T05
-status: todo
+status: done
 priority: medium
 state_hub_task_id: "4b59d4f4-d067-470b-b6c3-30c64cbc1010"
+started: "2026-05-28 ralph iter 5 (optional spike)"
+completed: "2026-05-28 ralph iter 5"
 ```
 
 **Description**: If time and review allow, implement a thin but real end-to-end slice of Profile 1 on top of Profile 0: new `kind="reflection"` support (or reuse retrospection), enhancement to `cya retrospect` (or new `cya reflect`) to capture verbal lessons, preferential activation for reflection kinds, rendering in `--explain-context` and final output, tests, and docs. All changes must be small, inspectable, and preserve safety invariants.