iso14971-risk-scorer
ISO 14971 risk estimation and scoring for medical devices. Severity × Probability matrix, residual risk evaluation.
- License 许可
- AGPL-3.0
- 最近更新
- 2026-11-14
- Stars
- ★ 25
- Forks
- 9
01 前言 Front Matter
- lang
- ko
- name
- iso14971-risk-scorer
- owner
- meddevice-agent
- scope
- co-safety
- mirror
- false
- status
- active
- created
- 2026-06-18
- version
- 1.1.0
- metadata
- { "type": "domain", "triggers": [ "ISO 14971", "위해 추정", "risk estimation", "심각도 발생확률 매트릭스", "severity probability matrix", "잔여위험", "residual risk", "ALARP" ], "legal_basis": [ "의료기기법 제28조 (의료기기 품질관리 — KGMP-MD)", "의료기기법 제31조 (의료기기 위해사항 보고)", "의료기기법 제6조 (의료기기 제조업 허가)", "의료기기 제조 및 품질관리 기준 (MFDS 고시 — KGMP-MD)", "ISO 14971:2019" ] }
- description
- ISO 14971 risk estimation and scoring for medical devices. Severity × Probability matrix, residual risk evaluation.
- lang_reason
- legal
- last_updated
- 2026-08-26
03 SKILL.md 正文
audit_exception: safety-os-skill-structure — Safety OS skills use the legal_basis-gated SSOT skill format (validated by scripts/skill-lifecycle-audit.ts and scripts/validate-skills.ts), not the generic template 5-section/7-frontmatter schema
ISO 14971 Risk Scorer Skill
Overview
ISO 14971 위해 추정 및 평가 — 심각도(Severity) × 발생확률(Probability) 매트릭스 기반 risk scoring.
Risk Estimation Matrix
| Severity \ Probability | Improbable | Remote | Occasional | Probable | Frequent |
|---|---|---|---|---|---|
| Negligible | Low | Low | Low | Medium | Medium |
| Minor | Low | Low | Medium | Medium | High |
| Serious | Low | Medium | Medium | High | Critical |
| Critical | Medium | Medium | High | Critical | Critical |
| Catastrophic | Medium | High | Critical | Critical | Critical |
Appendix — Cross-Walk to EHS 1-25 Scale (Consult-Mode Alignment Guidance)
Consult-mode alignment guidance only. This cross-walk exists so that device-risk outputs can be compared against the common EHS 1-25 scale used by
risk-assessment-agent(normative bands defined onrisk_score_beforeinevidence-models/domains/functional/risk-assessment/risk-assessment-record.json: 1-5 Low / 6-12 Medium / 13-19 High / 20-25 Critical; escalation threshold >= 13). It never replaces the qualitative matrix above — normative device-risk decisions stay on the ISO 14971 labels. Derivation rule: corner anchors Negligible×Improbable = 1 and Catastrophic×Frequent = 25; scale kept symmetric (cell(S,P) = cell(P,S)); monotone non-decreasing along both axes; every cell snapped into the band already published in the qualitative matrix, so band(numeric) always equals the ISO label.
| Severity \ Probability | Improbable | Remote | Occasional | Probable | Frequent |
|---|---|---|---|---|---|
| Negligible (S1) | 1 | 2 | 4 | 7 | 9 |
| Minor (S2) | 2 | 3 | 7 | 10 | 14 |
| Serious (S3) | 4 | 7 | 12 | 15 | 21 |
| Critical (S4) | 7 | 10 | 15 | 20 | 23 |
| Catastrophic (S5) | 9 | 14 | 21 | 23 | 25 |
Properties of this mapping:
- Band-exact: each numeric score lands in the same EHS band as its qualitative label in the matrix above (e.g. Serious+Probable = 15 = High), so consult-mode comparison can never contradict the ISO determination.
- Escalation-consistent: the minimum High cell is 14 and the maximum Medium cell is 12, so the EHS escalation threshold (score >= 13) partitions the matrix exactly into the union of High + Critical cells.
- Endpoints anchored: Catastrophic+Frequent = 25, Negligible+Improbable = 1.
ISO 14971 Process
Step 1: Hazard Identification
- Intended use / intended users
- Foreseeable misuse
- Reasonably foreseeable misuse
- Hazard identification (energy, biological, chemical, information)
Step 2: Risk Estimation
- Severity (S1-S5): Negligible → Catastrophic
- Probability (P1-P5): Improbable → Frequent
- Risk level = f(S, P) — matrix lookup
Step 3: Risk Evaluation
- Compare to risk acceptability criteria
- Broadly acceptable / ALARP / Unacceptable
Step 4: Risk Control
- Inherent safety by design
- Protective measures (alarms, interlocks)
- Information for safety (IFU, labeling)
- Residual risk disclosure
Step 5: Residual Risk Evaluation
- Overall residual risk
- Benefit-risk comparison (Class 3/4 devices)
- Risk management report
Korean Class-Specific Requirements
| Class | Risk Management Depth |
|---|---|
| Class 1 | Simplified (basic ISO 14971) |
| Class 2 | Standard ISO 14971 |
| Class 3 | Full + benefit-risk analysis |
| Class 4 | Full + clinical data + PMCF |
Output
Returns risk estimation:
{
"hazard_id": "H-001",
"hazard_description": "Electrical shock from power supply",
"severity": "Critical (S4)",
"probability": "Remote (P2)",
"risk_level_before": "Medium",
"risk_control": "Double insulation + grounding + IFU warning",
"residual_severity": "Critical (S4)",
"residual_probability": "Improbable (P1)",
"risk_level_after": "Low",
"residual_risk_acceptable": true,
"overall_residual_risk": "ALARP"
}
Legal Disclaimer
자동화 위해 추정 보조. 최종 위해 수용성 판정은 자격을 갖춘 의료기기 위해관리자 + 임상 평가자.
04 同仓库更多 Skills
i18n-layout
Text layout and encoding guidance: character encoding (UTF-8, legacy Korean code pages, BOM hazards), line endings, RTL/bidi handling, and script-specific font selection (Hangul focus). Use when: encoding is corrupted or mojibake appears, an RTL locale must render correctly, fonts must be chosen for a script, or CRLF/BOM issues surface.
i18n-layout
Text layout and encoding guidance: character encoding (UTF-8, legacy Korean code pages, BOM hazards), line endings, RTL/bidi handling, and script-specific font selection (Hangul focus). Use when: encoding is corrupted or mojibake appears, an RTL locale must render correctly, fonts must be chosen for a script, or CRLF/BOM issues surface.
i18n-layout
Text layout and encoding guidance: character encoding (UTF-8, legacy Korean code pages, BOM hazards), line endings, RTL/bidi handling, and script-specific font selection (Hangul focus). Use when: encoding is corrupted or mojibake appears, an RTL locale must render correctly, fonts must be chosen for a script, or CRLF/BOM issues surface.
i18n-layout
Text layout and encoding guidance: character encoding (UTF-8, legacy Korean code pages, BOM hazards), line endings, RTL/bidi handling, and script-specific font selection (Hangul focus). Use when: encoding is corrupted or mojibake appears, an RTL locale must render correctly, fonts must be chosen for a script, or CRLF/BOM issues surface.
i18n-layout
Text layout and encoding guidance: character encoding (UTF-8, legacy Korean code pages, BOM hazards), line endings, RTL/bidi handling, and script-specific font selection (Hangul focus). Use when: encoding is corrupted or mojibake appears, an RTL locale must render correctly, fonts must be chosen for a script, or CRLF/BOM issues surface.
i18n-layout
Text layout and encoding guidance: character encoding (UTF-8, legacy Korean code pages, BOM hazards), line endings, RTL/bidi handling, and script-specific font selection (Hangul focus). Use when: encoding is corrupted or mojibake appears, an RTL locale must render correctly, fonts must be chosen for a script, or CRLF/BOM issues surface.