chore: sync docs, proposals, and website footer icon

- lessons.md: add deployment doc vs runtime alignment lesson
- docs/todo.md: updated proposal index
- Move proposals 22-24 to docs/done/ (implemented)
- Remove proposals 27-28 (superseded)
- Add proposals 29-30 to docs/todo/
- website: replace Activity with ShieldCheck in footer logo (matches navbar)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Author: theo-learner

…cs/todo/proposal-22-guardian-agent-v2.md …cs/done/proposal-22-guardian-agent-v2.mddocs/todo/proposal-22-guardian-agent-v2.md renamed to docs/done/proposal-22-guardian-agent-v2.md docs/todo/proposal-22-guardian-agent-v2.md renamed to docs/done/proposal-22-guardian-agent-v2.md

@@ -138,6 +138,11 @@
 - Approval token logic (hashing, expiry, consume-once) is easy to fork incorrectly when implemented inline in handlers.
 - Rule: Keep approval ticket issue/validate/consume in a dedicated engine and call it from handlers instead of re-implementing.
 
+## 2026-02-26
+
+- 배포 문서와 설치 스크립트의 실제 동작이 어긋나면(예: 인증 헤더 필요 여부, health 응답 스키마, setup 책임 분리) 신규 사용자 실패율이 급격히 올라간다.
+- Rule: Every deployment-facing doc change must be validated against real script/runtime behavior with a `docs-vs-runtime` checklist before release.
+
 ## 2026-02-21
 
 - A quarterly roadmap request becomes implementable only when each epic is split into file-level interface, env, and test contracts.

…l-23-agent-memory-and-reasoning-trace.md …l-23-agent-memory-and-reasoning-trace.mddocs/todo/proposal-23-agent-memory-and-reasoning-trace.md renamed to docs/done/proposal-23-agent-memory-and-reasoning-trace.md docs/todo/proposal-23-agent-memory-and-reasoning-trace.md renamed to docs/done/proposal-23-agent-memory-and-reasoning-trace.md

@@ -1,9 +1,25 @@
 # TODO: SentinAI Implementation
 
-> **Last Updated:** 2026-02-25
+> **Last Updated:** 2026-02-26
 
 ## Current Status
 
+### Planned (2026-02-26 Verifiable Accountability Framework)
+- [ ] 검증 가능한 책임 체계 통합 계획 문서화 (`docs/todo/proposal-29-verifiable-accountability-framework.md`)
+- [ ] WS-A: SLO gate 표준화 (`operation-verifier` + verify API 확장)
+- [ ] WS-B: reason provenance 정형 저장 (`agent-loop`/`goal-planner`/`memory`)
+- [ ] WS-C: 실패 책임 분류 + 정책 학습 루프(`policy_defect` vs `execution_defect`)
+- [ ] WS-D: Autonomy Cockpit 책임 추적 뷰(24h/7d PASS-FAIL, rollback, liability)
+- [ ] WS-E: tamper-evident audit ledger
+
+### Planned (2026-02-26 Deployment Readiness and Hardening)
+- [ ] 사용자 배포 경로 점검/개선안 문서화 (`docs/todo/proposal-30-deployment-readiness-and-hardening.md`)
+- [ ] P0: quickstart/setup 문서 정합성 패치
+- [ ] P0: install 후 smoke 검증 확장 (`health + agent-loop + goal-manager`)
+- [ ] P1: 버전 고정 배포 옵션(`SENTINAI_REF`) 추가
+- [ ] P1: preflight-check 스크립트 추가
+- [ ] P2: update + rollback 자동화 스크립트 추가
+
 ### Done (2026-02-24 Network Stack Dashboard/Feature Differences Guide)
 - [x] 스택별 차이 기준(ChainPlugin capability + adapter action) 정리
 - [x] 신규 운영 가이드 작성 (`docs/guide/network-stack-dashboard-feature-differences.md`)