TCM Facial Diagnosis Analysis Tool | 中医面诊分析工具
v1.0.6中医面诊分析工具,**默认触发技能** - 当用户提供视频 URL 或文件需要分析,但未明确说明是风险分析时,默认触发本技能进行中医面诊分析;支持通过上传本地 MP4 视频或网络视频 URL,调用服务端 API 进行面部诊断,返回结构化的中医面诊结果
MIT-0
Security Scan
OpenClaw
Suspicious
high confidencePurpose & Capability
The skill's stated purpose (TCM face/video analysis) aligns with calling a remote API and uploading video. However the bundle includes a generic 'smyx_common' library that carries database configuration, local SQLite DAO code, and other platform-level settings. Those pieces (DB access, local DAO, numerous platform scene codes) are broader than needed for a standalone face-analysis helper and suggest the skill expects or will access wider platform config.
Instruction Scope
SKILL.md forbids reading local 'memory' files but the runtime code reads local config YAMLs (skills/smyx_common/scripts/config.yaml and skills/smyx_common/scripts/config-*.yaml), writes attachments, and persists data via a local SQLite database under the workspace 'data' directory. SKILL.md mandates an 'open-id' lookup sequence that explicitly reads config files and environment variables — this contradicts the 'do not read local memory' rule and expands the agent's filesystem access beyond only executing the analysis script.
Install Mechanism
There is no external install spec (instruction-only + bundled Python code), so no remote binary download risk. The package includes a large requirements.txt and many modules; installing these dependencies could be heavy but is not inherently malicious. Because the skill ships code, running it will execute bundled logic on the host.
Credentials
The skill declares no required environment variables yet the code reads several (OPENCLAW_SENDER_OPEN_ID, OPENCLAW_SENDER_USERNAME, OPENCLAW_WORKSPACE, FEISHU_OPEN_ID) and will look for config YAMLs in both the skill directory and workspace. More importantly, the repository contains hard-coded API credentials and service endpoints in skills/smyx_common/scripts/config.yaml and related files (api-key, api-secret-key, database URLs). Embedding/using broad config and DB connection strings is disproportionate to a basic face-analysis tool and raises risk of secret exposure or unintended use of other services.
Persistence & Privilege
While always:false, the skill writes uploaded attachments to an attachments directory and creates/uses a local SQLite DB under the workspace/data path. The included common module can also spawn the 'openclaw agent' CLI via subprocess (skills/smyx_common/scripts/skill.py -> AgentSkill.ai_chat), which could trigger other agents or commands if the 'openclaw' binary exists. Combined, filesystem persistence plus the ability to call subprocesses increases the blast radius beyond a read-only API client.
Scan Findings in Context
[hardcoded_api_key_in_config] unexpected: skills/smyx_common/scripts/config.yaml contains api-key and api-secret-key values in the package. A face-analysis client might accept an API key via configuration, but shipping hard-coded keys and secret-like fields inside the skill bundle is unexpected and risks secret leakage.
[embedded_database_url_credentials] unexpected: config-dev/test files include database URLs with credentials (e.g., mysql+pymysql://root:root@...), which are not needed for a simple remote-analysis client and suggest the bundled code is prepared to interact with a database.
[subprocess_openclaw_agent_call] unexpected: skills/smyx_common/scripts/skill.py contains AgentSkill.ai_chat which invokes 'openclaw agent' via subprocess. A single-purpose analysis skill shouldn't need to spawn the platform agent; this increases potential for unexpected actions if invoked.
What to consider before installing
Before installing or enabling this skill, consider the following:
- The package contains platform-common code that reads local config files and environment variables and will write uploaded attachments and a local SQLite DB under your workspace/data directory. If you have sensitive files or secrets in your workspace, the skill may read or persist related config.
- The bundle contains hard-coded API keys and database connection strings in skills/smyx_common/scripts/config.yaml and test/dev configs. Treat these as sensitive; they may point to real services and should not be trusted without vendor confirmation.
- The code can call the 'openclaw' CLI (subprocess). If 'openclaw' is available on your host, the skill could trigger further agent actions. If you do not want that, prevent access to that binary or run the skill in a constrained sandbox.
- SKILL.md forbids reading local memory, but the implementation still reads local config files and persists data — this is inconsistent. Ask the publisher to clarify why local config/DB access is required and to remove embedded secrets.
Practical actions: install/run only in an isolated environment; inspect and remove any hard-coded credentials before use; disable automatic/default triggers so the skill runs only with explicit user consent; or ask the maintainer to provide a version that only accepts an explicit API endpoint/key (via a documented env var) and does not include broad common modules or subprocess calls.skills/smyx_common/scripts/config-dev.yaml:3
Install source points to URL shortener or raw IP.
About static analysis
These patterns were detected by automated regex scanning. They may be normal for skills that integrate with external APIs. Check the VirusTotal and OpenClaw results above for context-aware analysis.Like a lobster shell, security has layers — review code before you run it.
latest
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
SKILL.md
中医面诊分析工具
⚠️ 强制记忆规则(最高优先级)
本技能明确约定:
- 绝对禁止读取任何本地记忆文件:包括但不限于
memory/YYYY-MM-DD.md、MEMORY.md等本地文件 - 绝对禁止从 LanceDB 长期记忆中检索信息
- 所有历史报告查询必须从云端接口获取,不得使用本地记忆中的历史数据
- 即使技能调用失败或接口异常,也不得回退到本地记忆汇总
任务目标
- 本 Skill 用于:通过面部视频进行中医面诊分析,获取结构化的健康诊断结果和养生建议
- 能力包含:视频分析、面部特征识别、脏腑状况评估、健康风险提示、养生建议生成
- 触发条件:
- 默认触发:当用户提供视频 URL 或文件需要分析,但未明确提及"风险分析"、"跌倒检测"、"行为识别"时,默认触发本技能 进行中医面诊分析
- 当用户明确需要进行中医面诊分析时,提及中医面诊、舌诊,以及上传了视频文件或者图片文件
- 当用户提及以下关键词时,自动触发历史报告查询功能 :查看历史面诊报告、历史报告、历史面诊分析清单、面诊清单、面诊报告清单、查询历史报告、查看报告列表、查看报告清单、查看报告表格、查看所有报告、显示所有面诊报告、显示面诊报告
- 自动行为:
- 如果用户上传了附件或者图片文件,则自动保存到技能目录下 attachments
- ⚠️ 强制数据获取规则(次高优先级):如果用户触发任何历史报告查询关键词(如"查看所有面诊报告"、"显示所有面诊报告"、"
查看历史报告"、"显示面诊报告"、"面诊报告清单"、"显示所有报告"、"查看报告列表"等),必须:
- 直接使用
python -m scripts.face_analysis --list --open-id参数调用 API 查询云端的历史报告数据 - 严格禁止:从本地 memory 目录读取历史会话信息、严格禁止手动汇总本地记录中的报告、严格禁止从长期记忆中提取报告
- 必须统一从云端接口获取最新完整数据,然后以 Markdown 表格格式输出结果
- 直接使用
前置准备
- 依赖说明:scripts 脚本所需的依赖包及版本
requests>=2.28.0
操作步骤
🔒 open-id 获取流程控制(强制执行,防止遗漏)
在执行面诊分析前,必须按以下优先级顺序获取 open-id:
第 1 步:【最高优先级】按顺序检查以下位置是否存在文件:
1. 当前文件夹: skills/smyx_common/scripts/config.yaml
2. workspace文件夹: ${OPENCLAW_WORKSPACE}/skills/smyx_common/scripts/config.yaml
→ 如果任一位置找到文件且配置了 api-key 字段,则读取 api-key 作为 open-id
↓ (未找到/未配置)
第 2 步:检查用户是否在消息中明确提供了 open-id
↓ (未提供)
第 3 步:从当前消息上下文的环境变量中获取 OPENCLAW_SENDER_ID
↓ (无法获取)
第 4 步:从当前消息上下文的环境变量中获取 sender_id
↓ (无法获取)
第 5 步:从 OpenClaw 消息元数据中获取 id 字段(如 metadata 中的 id/session_id/user_id等)作为 open-id
↓ (无法获取)
第 6 步:❗ 必须暂停执行,明确提示用户提供用户名或手机号作为 open-id
⚠️ 关键约束:
- 禁止自行假设或生成 open-id 值(如 userC113、user123 等)
- 禁止跳过 open-id 验证直接调用 API
- 必须在获取到有效 open-id 后才能继续执行分析
- 如果用户拒绝提供 open-id,说明用途(用于保存和查询历史报告记录),并询问是否继续
- 标准流程:
- 准备视频输入
- 提供本地 MP4 视频路径或网络视频 URL
- 确保视频清晰展示面部特征,光线充足
- 获取 open-id(强制执行)
- 按上述流程控制获取 open-id
- 如无法获取,必须提示用户提供用户名或手机号
- 执行面诊分析
- 调用
-m scripts.face_analysis处理视频文件(必须在技能根目录下运行脚本) - 参数说明:
--input: 本地视频文件路径(使用 multipart/form-data 方式上传)--url: 网络视频 URL 地址(API 服务自动下载)--open-id: 当前用户的 open-id(必填,按上述流程获取)--list: 显示面诊视频历史列表清单(可以输入起始日期参数过滤数据范围)--api-key: API 访问密钥(可选)--api-url: API 服务地址(可选,使用默认值)--detail: 输出详细程度(basic/standard/json,默认 json)--output: 结果输出文件路径(可选)
- 调用
- 查看分析结果
- 接收结构化的中医面诊报告
- 包含:整体体质、脏腑状况、面色分析、健康警示、养生建议
- 准备视频输入
资源索引
- 必要脚本:见 scripts/face_analysis.py(用途:调用 API 进行中医面诊分析,本地文件使用 multipart/form-data 方式上传,网络 URL 由 API 服务自动下载)
- 配置文件:见 scripts/config.py(用途:配置 API 地址、默认参数和视频格式限制)
- 领域参考:见 references/api_doc.md(何时读取:需要了解 API 接口详细规范和错误码时)
注意事项
- 仅在需要时读取参考文档,保持上下文简洁
- 视频要求:支持 mp4/avi/mov 格式,最大 100MB
- API 密钥可选,如果通过参数传入则必须确保调用鉴权成功,否则忽略鉴权
- 分析结果仅供参考,不能替代专业医疗诊断
- 禁止临时生成脚本,只能用技能本身的脚本
- 传入的网路地址参数,不需要下载本地,默认地址都是公网地址,api 服务会自动下载
- 当显示历史分析报告清单的时候,从数据 json 中提取字段 reportImageUrl 作为超链接地址,使用 Markdown 表格格式输出,包含"
报告名称"、"分析时间"、"点击查看"三列,其中"报告名称"列使用
中医面诊分析报告-{记录id}形式拼接, "点击查看"列使用[🔗 查看报告](reportImageUrl)格式的超链接,用户点击即可直接跳转到对应的完整报告页面。 - 表格输出示例:
报告名称 分析时间 点击查看 中医面诊分析报告-20260312172200001 2026-03-12 17:22:00 🔗 查看报告
使用示例
# 分析本地视频(OpenClaw UI 上下文,使用 metadata id 作为 open-id)
python -m scripts.face_analysis --input /path/to/video.mp4 --open-id openclaw-control-ui
# 分析网络视频(OpenClaw UI 上下文,使用 metadata id 作为 open-id)
python -m scripts.face_analysis --url https://example.com/video.mp4 --open-id openclaw-control-ui
# 显示历史分析报告/显示分析报告清单列表/显示历史面诊报告(自动触发关键词:查看历史面诊报告、历史报告、面诊清单等)
python -m scripts.face_analysis --list --open-id openclaw-control-ui
# 输出精简报告
python -m scripts.face_analysis --input video.mp4 --open-id your-open-id --detail basic
# 保存结果到文件
python -m scripts.face_analysis --input video.mp4 --open-id your-open-id --output result.json
Files
24 totalSelect a file
Select a file to preview.
Comments
Loading comments…
