Sql To Doc

Security checks across malware telemetry and agentic risk

Overview

The skill is coherent, but it can run broad SQL queries and publish results into Feishu documents without clear safeguards around sensitive data, destination visibility, or confirmation.

Install only for users who are allowed to query the relevant DataWorks/MaxCompute data and publish it to Feishu. Use read-only, least-privileged credentials, confirm the exact SQL, destination, document visibility, and notification recipient before each run, and avoid exporting sensitive data unless your organization has approved that workflow.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill explicitly creates Feishu cloud documents and may send notifications, but the description/output section does not clearly warn users that it performs external write actions and messaging side effects. This can cause users to unintentionally publish query results to shared locations or notify unintended recipients, especially when the SQL output may contain sensitive business data.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill encourages arbitrary SQL execution and export of results into a document without a prominent privacy or data-sensitivity warning. In this context, that omission is dangerous because users may query confidential datasets and automatically copy the results into Feishu, increasing the risk of data exfiltration, over-sharing, or persistence of sensitive information in collaboration systems.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal