Biomarker Investigation Zhcn
PassAudited by VirusTotal on May 13, 2026.
Overview
Type: OpenClaw Skill Name: biomarker-investigation-zhcn Version: 1.0.4 The skill bundle is a legitimate tool designed for life sciences research, specifically focused on biomarkers and patent analysis using PatSnap's services. It provides clear instructions for setting up Model Context Protocol (MCP) connections to PatSnap endpoints (e.g., connect.patsnap.com) and includes a robust self-check mechanism to ensure connectivity before execution. There is no evidence of data exfiltration, malicious execution, or harmful prompt injection; the instructions are strictly aligned with the stated purpose of academic and patent investigation.
Findings (0)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Your PatSnap API key may be stored in the MCP configuration and used for service access, billing, quotas, or account-linked requests.
The skill requires a PatSnap API key and places it in the MCP connection URL. This is expected for the PatSnap integration, but it is still a sensitive account credential.
登录 https://open.patsnap.com ，进入 **API Keys**，创建一个新 Key。 ... "https://connect.patsnap.com/096456/logic-mcp?apiKey=sk-xxxxxxxxxxxx"
Use a dedicated, revocable PatSnap key with the least scope available, avoid sharing MCP config or logs containing the URL, and rotate the key if it is exposed.
Biomarker, disease, drug, patent, or clinical-trial queries may leave the local agent environment and be processed by PatSnap's service.
The skill routes research queries through a remote PatSnap MCP service. This is central to the skill's purpose, but it means query terms and tool requests are sent to an external provider.
claude mcp add --transport http pharma_intelligence "https://connect.patsnap.com/096456/logic-mcp?apiKey=sk-xxxxxxxxxxxx" ... **重要提示**：优先使用生命科学 MCP 服务进行数据检索
Verify the PatSnap endpoint and account terms before use, and avoid entering confidential or proprietary research questions unless your organization approves that data sharing.
The agent may contact PatSnap even before answering a biomarker question and may make many read-only fetch requests during a research task, potentially affecting quotas or logs.
The skill instructs the agent to make an automatic MCP connectivity test and to perform bulk detail fetches for search results under a threshold. These are read-only and purpose-aligned, but they can increase remote calls.
每次技能加载后、处理任何用户查询之前，必须先执行以下自检 ... 使用 `ls_target_fetch` 按名称查询 EGFR ... 若 `_search` 工具返回结果不超过 100 条 ... 必须使用全部搜索结果 ID 调用 `_fetch`
Keep tool-call visibility or approvals enabled if quota, cost, or auditability matters, and stop the task if the agent begins making broader calls than needed.