← 返回 Skills 市场
114
总下载
0
收藏
1
当前安装
2
版本数
在 OpenClaw 中安装
/install sfe-data-viewer
功能描述
SFE数据查询与分析工具，用于查询医药销售效能数据
安全使用建议
This skill appears to actually provide SFE query scripts and API docs, but it contains a risky runtime step: if the auth helper (cms-auth-skills) is missing the SKILL.md tells the agent to run npx/clawhub to install it — including a fallback GitHub URL. Before installing or enabling this skill:
- Verify the source of cms-auth-skills (review its repository and verify maintainers) rather than allowing automatic npx/GitHub install.
- Ask the skill author to declare required binaries (python, node/npx, clawhub) and to provide a formal install spec with trusted release URLs or checksums.
- Review the bundled Python scripts (scripts/) yourself or in an isolated environment to ensure they only call the documented erp-web.mediportal.com.cn endpoints and do not exfiltrate data elsewhere.
- Confirm how the appKey is retrieved/stored by cms-auth-skills (avoid trusting opaque credential-handling code). Consider providing credentials manually instead of allowing remote installers to obtain/store them.
- If you must run it, do so first in an isolated environment (sandbox/container) and do not run the npx install/fallback GitHub command until you’ve inspected that code.
Because of the undeclared remote-install instructions and metadata mismatches, proceed cautiously and get the above clarifications or code review before trusting this skill with sensitive data.
功能分析
Type: OpenClaw Skill
Name: sfe-data-viewer
Version: 1.0.1
The skill bundle is a legitimate data retrieval tool for Sales Force Effectiveness (SFE) in the pharmaceutical industry. It contains Python scripts that interface with APIs at mediportal.com.cn to query zone, product, and project data. While the scripts exhibit a security vulnerability by disabling SSL verification (verify=False in multiple files like scripts/sfe-user/zone.py), this appears to be a common development oversight rather than intentional malice. The bundle includes a specialized 'TOON' encoder (scripts/common/toon_encoder.py) to optimize LLM token consumption, and its instructions in SKILL.md are consistent with standard agent-based task automation and dependency management.
能力评估
Purpose & Capability
The skill claims to be a read-only SFE data query/analysis tool and includes many Python scripts and OpenAPI docs that align with that purpose. However, the SKILL.md requires installing and using 'cms-auth-skills' (for appKey handling) via npx/clawhub if missing — yet the package metadata declares no required binaries (node/npx, python) or install steps. The scripts are Python but Python is not declared as a required binary. This mismatch between claimed requirements and actual runtime steps is an incoherence.
Instruction Scope
The runtime instructions mandate executing bundled Python scripts (not calling APIs directly) and say to read cms-auth-skills/SKILL.md for authentication. They also allow reading user files/URLs to build inputs. Critically, they instruct to run npx/clawhub to install cms-auth-skills from npm or GitHub if not present — a remote-install step that pulls and runs code outside the declared package. The instructions thus reach beyond simple API-doc reading into executing remote installs and local scripts, which increases risk.
Install Mechanism
There is no formal install spec, but SKILL.md instructs use of 'npx clawhub@latest install cms-auth-skills --force' and a fallback GitHub URL. Installing code via npx or from an arbitrary GitHub repo is higher risk (arbitrary code download/run). The skill does not declare Node/npx as required, nor provide hashes or trusted release URLs for dependencies, making the remote-install guidance disproportionate and potentially unsafe.
Credentials
The APIs documented require an 'appKey' header; the skill delegates appKey preparation to cms-auth-skills rather than listing env vars itself. No environment variables or secrets are declared in metadata, which is coherent in one sense, but the dependency on an external auth skill means secret handling is delegated and not visible here. That lack of explicit credential declaration is a transparency concern (you can't easily audit how the appKey is obtained/stored).
Persistence & Privilege
The skill does not request always:true, does not declare system config paths or permanent privileges, and is user-invocable only. There is no automatic forced installation listed. Autonomous invocation is allowed (platform default) but not combined with other high-privilege flags.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install sfe-data-viewer - 安装完成后，直接呼叫该 Skill 的名称或使用
/sfe-data-viewer触发 - 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
v1.0.1
更新技能描述和文档
v1.0.0
首次发布：SFE数据查询与分析工具，用于查询医药销售效能数据
元数据
常见问题
Sfe Data Viewer 是什么？
SFE数据查询与分析工具，用于查询医药销售效能数据. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 114 次。
如何安装 Sfe Data Viewer？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install sfe-data-viewer」即可一键安装，无需额外配置。
Sfe Data Viewer 是免费的吗？
是的，Sfe Data Viewer 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Sfe Data Viewer 支持哪些平台？
Sfe Data Viewer 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 Sfe Data Viewer？
由 spzwin（@spzwin）开发并维护，当前版本 v1.0.1。
推荐 Skills