c24165b8bc
에이전트 8종(대화형/프로그래머 C·S/디자이너/설계자/기획자/QA/PD/PM)에게 [공통 기본 능력 + 롤별 Specialty] 2층으로 지식을 주입하기 위한 재분류. 문서 내용·포맷은 무수정, 폴더 이동만 (6,372개 문서 수 보존 확인). - Topic_Programming → Domain_Programming (내부 구조 보존) - Topic_Graphic → Domain_Design - Topic_Business → Domain_Product - Topic_General → Domain_General - _Common 신설: Math(구 Topic_Math_Specialty), Reasoning(구 General/From_Thinking & Reasoning), Reasoning_Creativity(구 General/From_창의성), Communication(Poetic_Blog_Writing + From_writing) - 타 도메인의 From_* 폴더는 유지 (출처 표기일 뿐, 이미 도메인에 맞게 분류된 문서) - 빈 폴더 정리 (memory/procedures) - 에이전트→폴더 매핑은 workspace의 .astra/agent-knowledge-map.json (9개 에이전트) Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
4.9 KiB
4.9 KiB
id, title, category, status, canonical_id, aliases, duplicate_of, source_trust_level, confidence_score, verification_status, tags, raw_sources, last_reinforced, github_commit, tech_stack
| id | title | category | status | canonical_id | aliases | duplicate_of | source_trust_level | confidence_score | verification_status | tags | raw_sources | last_reinforced | github_commit | tech_stack | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| wiki-2026-0508-tara | TARA (Threat Analysis and Risk Assessment) | 10_Wiki/Topics | verified | self |
|
none | A | 0.9 | applied |
|
2026-05-10 | pending |
|
TARA (Threat Analysis and Risk Assessment)
매 한 줄
"매 automotive cybersecurity 의 mandatory threat modeling". 매 ISO/SAE 21434 의 core process — 매 vehicle E/E system 의 cybersecurity risk 의 systematic 의 identify + assess + treat. 2026 의 UNECE R155 type-approval 의 prerequisite.
매 핵심
매 7 steps (ISO 21434 §15)
- Asset identification: 매 cybersecurity property (CIAA) 의 정의.
- Threat scenario identification: STRIDE/EVITA 의 적용.
- Impact rating: Safety, Financial, Operational, Privacy (SFOP).
- Attack path analysis: 매 attack tree / kill chain.
- Attack feasibility rating: time, expertise, knowledge, opportunity, equipment.
- Risk determination: Impact × Feasibility → 1-5 risk value.
- Risk treatment: avoid, reduce, share, retain.
매 STRIDE for automotive
- Spoofing — 매 ECU identity forgery (CAN ID).
- Tampering — 매 firmware mod, OTA hijack.
- Repudiation — 매 audit log absence.
- Info disclosure — 매 GPS/PII leak.
- DoS — 매 CAN bus flood.
- EoP — 매 infotainment → drive ECU pivot.
매 vs IT threat modeling
- Lifecycle: 매 15 yr vehicle 의 — 매 long-term update.
- Safety coupling: 매 cyber → physical harm (브레이크).
- Supply chain: 매 multi-tier (OEM ← T1 ← T2).
- Standards: 매 ISO 21434 + UNECE R155 의 mandatory.
매 응용
- New vehicle development (concept phase).
- CSMS audit evidence.
- Post-incident re-assessment.
💻 패턴
Asset table (YAML)
assets:
- id: ASSET-01
name: Brake control message
properties: [integrity, availability]
rationale: Loss → safety-critical
- id: ASSET-02
name: GPS trajectory log
properties: [confidentiality]
rationale: PII / GDPR
- id: ASSET-03
name: ECU firmware
properties: [integrity, authenticity]
Damage scenario × impact rating
damage_scenarios:
- id: DS-01
asset: ASSET-01
description: Attacker injects fake brake-release message
impact:
safety: severe # S3 (life-threatening)
financial: moderate
operational: major
privacy: negligible
overall: severe
Threat scenario (STRIDE)
threats:
- id: TS-01
target: ASSET-01
stride: tampering
description: Spoof brake CAN frame from compromised infotainment
leads_to: DS-01
Attack path
Internet → Telematics gateway (CVE-X) → Internal CAN gw bypass →
Body CAN → Inject 0x100 brake frame
Feasibility rating (ISO 21434 Annex G)
attack_path: AP-01
factors:
elapsed_time: 1_to_6_months # 4 pts
expertise: expert # 6 pts
knowledge: restricted # 3 pts
window: easy # 1 pts
equipment: specialized # 4 pts
total: 18 # → Medium feasibility
Risk matrix
Feasibility →
Impact ↓ VeryLow Low Med High
Severe 2 3 4 5
Major 1 2 3 4
Moderate 1 1 2 3
Negligible 1 1 1 2
Treatment plan
risk: R-01 (severe × medium = 4)
treatment: reduce
controls:
- C-01: Message Authentication Code (MAC) on brake CAN
- C-02: Secure boot on telematics gateway
- C-03: Firewall rules between domains
residual_risk: 2 # acceptable
매 결정 기준
| 상황 | Approach |
|---|---|
| Automotive E/E (mandatory) | TARA per ISO 21434 |
| IT system | STRIDE / PASTA |
| Industrial control (ICS) | IEC 62443 ZCR |
| Generic risk | NIST 800-30 |
기본값: 매 ISO 21434 TARA — 매 OEM type approval (UNECE R155) 의 required.
🔗 Graph
🤖 LLM 활용
언제: threat scenario brainstorming, attack tree generation, control gap analysis. 언제 X: final risk decision (human cybersecurity engineer 의 sign-off).
❌ 안티패턴
- One-shot TARA: 매 lifecycle 의 update 의 — 매 vuln/incident 의 trigger.
- Generic STRIDE only: 매 automotive-specific 의 EVITA 의 augment.
- Skipping feasibility: 매 impact-only 의 risk 의 inflated.
- Doc theater: 매 controls 의 implement 의 X — 매 evidence 의 X.
🧪 검증 / 중복
- Verified (ISO/SAE 21434:2021, UNECE WP.29 R155, EVITA project).
- 신뢰도 A.
🕓 Changelog
| 날짜 | 변경 |
|---|---|
| 2026-05-08 | Phase 1 |
| 2026-05-10 | Manual cleanup — ISO 21434 7-step TARA, automotive STRIDE |