跳到正文

D002 数据结构与关联关系分析

分析医疗数据的字段、记录单位、主键质量与跨表关联,记录数据缺口及连接条件,为后续处理提供结构依据。

个人 Skill知识加工11 个文件浏览次数 0
开始使用

下载前阅读 SKILL.md 中的使用条件,并确认所用工具支持对应能力。

输入资料
Required: D001 source_manifest.json the exact local source file/directory that D001 inventoried Preferred invocation:D002 must verify that every profiled file resolves inside the declared input boundary.Supported determi… 阅读本节完整说明
交付内容
Always produce: schema_linkage_profile.json schema_linkage_report.md linkage_edges.csv Prefer:./runs/haonan-d002/<package-id>/The machine-readable contract is defined in assets/schema_linkage_profile.schema.json and expl… 阅读本节完整说明
  1. 获取完整包,保留配套文件的目录结构。
  2. 按照原文准备所需工具、依赖与输入资料。
  3. 阅读对应使用步骤与输入输出说明,在自己的环境中验证结果。

使用提示摘自原文,配套文件保留来源版本;实际运行条件以各包说明为准。

正文

获取原文件

references/security-policy.md

Threat model

D002 assumes the source package may contain PHI, credentials, malicious prompt text, scripts, malformed files, macro-enabled Office files, symlinks, or misleading schemas.

Hard boundaries

  • local-only
  • standard-library-only
  • source read-only
  • no input code execution
  • no shell execution
  • no remote DB connections
  • no arbitrary SQL from input
  • no archive expansion except controlled XLSX OpenXML inspection
  • no raw value output

Value-level profiling

D002 needs local values to validate uniqueness and joins. Identifier values are normalized and HMAC-tokenized with an ephemeral per-run secret. The secret and tokens are never serialized.

Direct-identifier/free-text columns receive restricted profiling. Do not create value distributions or examples.

SQLite

Only base tables are inspected. Database is opened read-only/query-only; extension loading is disabled. Views are not queried because their SQL may invoke unexpected functions or expensive expressions.

Output review

Generated JSON/Markdown/CSV must not contain known raw source values. Synthetic self-tests explicitly check this invariant.

CONTACT HAONAN

联系haonan

医疗 AI · 产品与独立开发

刘浩南

关注医疗 AI 从 demo 到真实落地,聚焦产品化、商业化与院内工作流效率提升。

12年医疗产研经验,大湾区头部医疗AI公司前产品总监,独立开发者

医疗 AI 产品化医疗 AI 商业化院内工作流效率
刘浩南的微信联系二维码
微信扫一扫,联系 haonan