← 返回 Skills 市场
gechengling

Insurance Claims Intelligence Expert

作者 gechengling · GitHub ↗ · v1.2.0 · MIT-0
cross-platform ✓ 安全检测通过
93
总下载
0
收藏
0
当前安装
3
版本数
在 OpenClaw 中安装
/install insurance-claims-intelligence
功能描述
提供多模态医疗票据OCR识别、智能判责、反欺诈检测和全险种覆盖的保险理赔智能分析与自动化支持。
使用说明 (SKILL.md)

\r \r

Insurance Claims Intelligence Expert / 保险行业智能理赔专家\r

\r

⚠️ DISCLAIMER / 免责声明\r

  • English: This skill provides advisory templates, checklists, and decision-support frameworks ONLY. It does NOT contain executable models, trained GNN weights, or production OCR integrations. All accuracy figures (e.g., "92%-96%") are literature-reported benchmarks or design targets, NOT validated results of this skill. ALL claim approvals, denials, payout amounts, and fraud labels MUST be reviewed and confirmed by a licensed insurance professional before use. This skill is NOT a substitute for human judgment or regulatory compliance review.\r
  • 中文: 本Skill仅提供咨询模板、检查清单和决策支持框架,不含可执行模型、已训练GNN权重或生产级OCR集成。所有准确率数据(如"92%-96%")均来自文献基准或设计目标,非本Skill实测结果。所有理赔核准、拒付、赔付金额及欺诈标签,必须经持证保险专业人士审核确认后方可使用。本Skill不可替代人工判断或监管合规审查。\r \r 🔒 DATA SECURITY / 数据安全\r
  • Medical invoices, diagnosis records, and claimant data are sensitive personal information under China's Personal Information Protection Law (PIPL). Before using OCR features, obtain user consent, redact/remove unnecessary PII, prefer on-prem/private deployment for production, and confirm the OCR vendor's data retention and cross-border transfer terms.\r
  • API keys and credentials MUST be stored in environment variables or a secret manager. Never hardcode keys in production systems.\r
  • English: This skill provides advisory templates, checklists, and decision-support frameworks ONLY. It does NOT contain executable models, trained GNN weights, or production OCR integrations. All accuracy figures (e.g., "92%-96%") are literature-reported benchmarks or design targets, NOT validated results of this skill. ALL claim approvals, denials, payout amounts, and fraud labels MUST be reviewed and confirmed by a licensed insurance professional before use. This skill is NOT a substitute for human judgment or regulatory compliance review.\r
  • 中文: 本Skill仅提供咨询模板、检查清单和决策支持框架,不含可执行模型、已训练GNN权重或生产级OCR集成。所有准确率数据(如"92%-96%")均来自文献基准或设计目标,非本Skill实测结果。所有理赔核准、拒付、赔付金额及欺诈标签,必须经持证保险专业人士审核确认后方可使用。本Skill不可替代人工判断或监管合规审查。\r \r ---\r \r

Artifact Type / 作品类型\r

\r This is a documentation-and-template skill. It contains:\r

  • ✅ Workflow checklists and decision trees\r
  • ✅ Report templates and output formats\r
  • ✅ Reference architectures and integration guidance\r
  • ✅ Example Python code (requires your own API keys and data)\r \r It does NOT contain:\r
  • ❌ Pre-trained ML/GNN models\r
  • ❌ Executable OCR or claims processing code\r
  • ❌ Bundled third-party API credentials\r \r ---\r \r

Trigger Keywords / 触发关键词\r

\r English Triggers: insurance claims advisory, claims workflow, claim analysis, medical OCR guidance, insurance fraud assessment, claim liability review, policy clause analysis, anti-fraud checklist, insurance tech reference, claims report template\r \r 中文触发词: 保险理赔咨询 / 理赔流程指导 / 理赔分析 / 医疗发票识别指导 / 理赔判责参考 / 责任认定流程 / 医疗险理赔 / 重疾险理赔 / 寿险理赔 / 意外险理赔 / 车险理赔 / 财产险理赔 / 理赔反欺诈 / 欺诈检测参考 / 骗保识别指导 / 理赔风控参考 / 保险条款解读 / 责任免除说明 / 保障范围分析 / 赔付比例计算 / 产品对比参考 / 条款比对指导 / 合同解读参考\r \r ---\r \r

Core Capabilities / 核心能力(咨询框架)\r

\r

1. Medical Receipt OCR — Guidance Framework / 医疗票据OCR识别(指导框架)\r

\r 支持的票据类型(覆盖全场景):\r \r | Receipt Type / 票据类型 | Extracted Fields / 识别内容 | Insurance Types / 适用险种 |\r |------------------------|-------------------|------------------|\r | 全国统一门诊发票 | 发票号、医院、金额、明细项目 | 医疗险、意外险 |\r | 全国统一住院发票 | 入院/出院日期、总金额、自费比例 | 医疗险、重疾险 |\r | 医疗费用明细清单 | 药品明细、检查项目、单价、数量 | 医疗险 |\r | 医保结算单 | 医保账户支付、自付金额、报销比例 | 医疗险 |\r | 出院小结 | 诊断、住院天数、治疗经过 | 重疾险、寿险 |\r | 病历首页 | 主要诊断、手术名称、ICD编码 | 重疾险 |\r | 检查报告单 | 影像报告、检验结果 | 重疾险 |\r | 费用结算单 | 分项金额、总计金额 | 财产险、责任险 |\r \r

⚠️ OCR Data Handling / OCR数据处理提醒\r

  • Only send necessary fields to OCR providers; redact/unnecessary PII beforehand.\r
  • Confirm the OCR vendor's data retention policy (Prefer: no storage / auto-delete within 24h).\r
  • For production use, prefer private on-prem OCR deployment to avoid third-party data transfer.\r
  • 中文: 仅发送必要字段至OCR服务商;事前脱敏/删除非必要个人信息;确认OCR厂商数据留存策略(优先:不留存/24小时内自动删除);生产环境优先使用私有化本地部署OCR,避免第三方数据传输。\r \r 参考技术架构(需自行集成):\r \r
原始图像\r
  ↓\r
图像预处理(去噪/倾斜校正/二值化)\r
  ↓\r
CNN特征提取(ResNet50/EfficientNet)—— 需自行训练或调用云服务API\r
  ↓\r
RNN序列建模(BiLSTM)+ Attention机制\r
  ↓\r
CRF层解码 → 结构化文本输出\r
  ↓\r
字段标准化 → JSON/表格结构化结果\r
```\r
\r
### 2. Liability Determination — Advisory Framework / 理赔判责引擎(咨询框架)\r
\r
**咨询级判责检查清单(需人工逐项确认):**\r
\r
```text\r
规则1:等待期检查(人工确认)\r
  └─ 出险日期 - 保单生效日 \x3C 等待期 → 建议拒付,需人工复核\r
\r
规则2:既往症筛查(人工确认)\r
  └─ 既往症库匹配 → 责任免除 → 建议拒付/比例赔付,需人工复核\r
\r
规则3:免赔额校验(人工确认)\r
  └─ 累计自付金额 \x3C 免赔额 → 建议暂不赔付,需人工复核\r
\r
规则4:就诊机构核查(人工确认)\r
  └─ 非二级及以上公立医院(需视条款)→ 提示确认,需人工复核\r
\r
规则5:险种责任匹配(人工确认)\r
  └─ 就诊科室/诊断是否符合条款保障范围 → 建议全额/比例/拒付,需人工复核\r
```\r
\r
> **⚠️ IMPORTANT / 重要提醒**\r
> The liability determination output is a **decision-support suggestion ONLY**. Final approval/denial MUST be made by an authorized human reviewer. This skill does NOT auto-approve any claim amount.\r
> **中文:** 判责输出**仅为决策支持建议**,最终核准/拒付**必须由授权人工审核员作出**。本Skill不对任何理赔金额进行自动审批。\r
\r
### 3. Anti-Fraud Assessment — Advisory Framework / 反欺诈评估(咨询框架)\r
\r
**反欺诈检查清单(咨询级):**\r
\r
```text\r
检查项1:就诊频率异常\r
  └─ 同一被保人短期内多次就诊 → 标记,建议人工调查\r
\r
检查项2:票据真实性验证\r
  └─ 发票号重复 / 医院不存在 / 金额异常 → 标记,建议人工调查\r
\r
检查项3:诊断与用药匹配性\r
  └─ 诊断与开具药品明显不符 → 标记,建议人工调查\r
\r
检查项4:关系网络异常\r
  └─ 同一医生/医院集中出现在多起理赔 → 标记,建议人工调查\r
```\r
\r
> **🔒 Anti-Fraud Data Governance / 反欺诈数据治理**\r
> - Retention limit / 留存期限:反欺诈图谱数据建议留存不超过 2 年,除非监管要求的更长留存期。\r
> - Access control / 访问控制:图谱查询权限仅开放给授权欺诈调查员,禁止非授权人员访问。\r
> - Data correction workflow / 数据更正流程:被保人有权请求更正错误数据,必须在 15 个工作日内处理。\r
> - Poisoning safeguard / 污染防护:新案件数据进入图谱前,须经人工审核确认,防止恶意污染。\r
\r
### 4. Claims Report Templates / 理赔报告模板\r
\r
```markdown\r
# 理赔分析报告(咨询草稿)\r
**生成时间**: YYYY-MM-DD HH:mm\r
**案件编号**: CL-XXXXXXXX\r
**险种类别**: [险种名称]\r
**处理状态**: [咨询草稿 — 需人工审核]\r
**免责声明**: 本报告为AI辅助生成的咨询草稿,所有结论须经持证理赔师审核确认后方可生效。\r
---\r
## 一、票据识别结果(仅供参考)\r
## 二、责任认定分析(仅供参考)\r
## 三、赔付计算参考(仅供参考)\r
## 四、反欺诈风险评估(仅供参考)\r
## 五、建议下一步行动(需人工确认)\r
```\r
\r
---\r
\r
## Compliance & Human Review / 合规与人工审核要求\r
\r
| Compliance Item / 合规项 | Regulatory Basis / 监管依据 | Human Review Requirement / 人工审核要求 |\r
|--------------------|--------------------|----------------------|\r
| 理赔时效 | 《保险法》第23条 | 核定结果须经人工确认后发出 |\r
| 材料完整性 | 理赔管理办法 | 缺失材料列表由人工最终确认 |\r
| 反欺诈合规 | 《反保险欺诈工作办法》2024 | 欺诈标记须经人工调查确认 |\r
| 数据安全 | 《个人信息保护法》 | 医疗数据脱敏处理须经人工检查 |\r
| 资金安全 | 反洗钱规定 | 大额理赔须人工复核 + 主管审批 |\r
\r
**ALL outputs of this skill are drafts requiring licensed professional review. / 本Skill所有输出均为草稿,须经持证专业人士审核。**\r
\r
---\r
\r
## Output Format / 输出格式规范\r
\r
All outputs must include the following disclaimer:\r
\r
```markdown\r
> ⚠️ **免责声明 / Disclaimer**\r
> 本输出为AI辅助咨询草稿,所有理赔决定、拒付结论、赔付金额及欺诈标签\r
> 须经【持证保险理赔师】审核确认后方可生效。\r
> This is an AI-assisted draft. All claim decisions must be reviewed by a\r
> licensed insurance adjuster before taking effect.\r
```\r
\r
---\r
\r
## References / 参考文件\r
\r
| File / 文件 | Content / 内容说明 |\r
|------|---------|\r
| `references/claims_ocr_tech.md` | OCR技术架构参考 + 4家服务商对比 + Python示例代码(需自行配置API Key) |\r
| `references/claims_liability_engine.md` | 判责规则参考 + 机器学习模型参考 + 3家公司实践参考 |\r
| `references/claims_report_templates.md` | 报告模板 + 7种险种通知书模板参考 |\r
\r
> **⚠️ Reference files contain example code only. You must:**\r
> - Provide your own API keys and store them in environment variables\r
> - Provide your own training data and models\r
> - Ensure human review of all outputs before use\r
> - **中文:** 参考文件仅含示例代码,您必须:自行提供API密钥并存入环境变量;自行准备训练数据和模型;确保所有输出经人工审核后方可使用。\r
安全使用建议
This appears safe to install as a documentation/template skill, not an automated claims engine. Before using it with real cases, confirm that medical data handling complies with your policies and law, redact unnecessary PII, use secure API-key storage, verify any OCR provider’s retention and transfer terms, and require licensed human review before any claim decision or customer notice is issued.
功能分析
Type: OpenClaw Skill Name: insurance-claims-intelligence Version: 1.2.0 The skill bundle is a documentation-heavy advisory tool for insurance claims processing, providing structured checklists, report templates, and reference Python code. It contains no malicious logic, data exfiltration attempts, or harmful prompt injections; instead, it includes extensive disclaimers in SKILL.md and README.md emphasizing the necessity of human review and data privacy compliance (PIPL). Technical references in references/claims_ocr_tech.md and references/claims_liability_engine.md provide safe, standard integration examples for OCR and liability logic, correctly advising the use of environment variables for credential management.
能力标签
requires-oauth-tokenrequires-sensitive-credentials
能力评估
Purpose & Capability
The skill is coherent as an advisory/template skill for OCR guidance, claim review, fraud checklists, and report drafting, but the domain is high-impact because drafts may discuss approvals, denials, payouts, and fraud labels.
Instruction Scope
The artifacts repeatedly require licensed human review and say outputs are advisory, but some templates include final-decision placeholders that users must not treat as automated claim decisions.
Install Mechanism
There is no install spec or executable code, which lowers runtime risk; however, the registry lists the source as unknown and no homepage, so content provenance is limited.
Credentials
Example integrations use user-supplied OCR/LLM API keys and may send medical invoice images to cloud providers; this is purpose-aligned and disclosed with consent/redaction guidance.
Persistence & Privilege
The supplied artifacts do not implement persistence, but the anti-fraud guidance discusses retaining graph data, so any implementation should strictly control retention, access, correction, and poisoning safeguards.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install insurance-claims-intelligence
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /insurance-claims-intelligence 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.2.0
v1.2.0 Security & Compliance Update: (1) Added prominent disclaimers (EN/CN) stating this is advisory-only with NO executable models; (2) All accuracy figures now labeled as literature benchmarks, NOT validated results; (3) Removed ALL auto-approval language (no more 'auto-approve' or amount thresholds); (4) Added data security notices (PII redaction, API key management, vendor data retention policy); (5) Added anti-fraud data governance section (retention limits, access control, correction workflow); (6) All outputs now clearly labeled as 'drafts requiring licensed professional review'. Files changed: SKILL.md, README.md, references/claims_ocr_tech.md, references/claims_liability_engine.md.
v1.1.0
v1.1.0 Bilingual optimization: English metadata + summaries; Bilingual README.md; SEO title optimization for international users; Keywords: insurance claims, intelligent claims, medical OCR.
v1.0.0
首个保险行业全流程智能理赔Skill!整合多模态医疗票据OCR识别(百度/腾讯云/阿里云/合合信息)、智能理赔判责引擎(规则引擎+ML双驱动)、反欺诈知识图谱(GNN图神经网络)、行业全险种覆盖(医疗/重疾/寿险/意外/车险/财产险/团险),基于平安111极速赔、中国人寿智能理赔、太保数字劳动力实验室最佳实践构建,支持端到端理赔Agent,OCR识别→判责→核赔→反欺诈→合规检查全自动处理
元数据
Slug insurance-claims-intelligence
版本 1.2.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 3
常见问题

Insurance Claims Intelligence Expert 是什么?

提供多模态医疗票据OCR识别、智能判责、反欺诈检测和全险种覆盖的保险理赔智能分析与自动化支持。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 93 次。

如何安装 Insurance Claims Intelligence Expert?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install insurance-claims-intelligence」即可一键安装,无需额外配置。

Insurance Claims Intelligence Expert 是免费的吗?

是的,Insurance Claims Intelligence Expert 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Insurance Claims Intelligence Expert 支持哪些平台?

Insurance Claims Intelligence Expert 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Insurance Claims Intelligence Expert?

由 gechengling(@gechengling)开发并维护,当前版本 v1.2.0。

💬 留言讨论