ClawHub
Back to skill

Security audit

ali-promotion-roi

Security checks across malware telemetry and agentic risk

Overview

This skill is a read-only Alibaba promotion ROI reporter that clearly discloses its database access, though users should notice it automatically approves DB credential use through sql-linker-cli.

Install this only in an environment where the sql-linker-cli database credential gate is allowed to be approved by this skill. It reads business order and promotion data from the configured MySQL database, so prefer month filters for narrower reports and confirm the two table permissions are appropriate.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Intent-Code Divergence

High
Confidence
98% confidence
Finding
The docstring on `connect_db` says credential approval must be explicitly granted, which communicates a gated or user-mediated access model. In practice, the function immediately calls `db.explicit_credential_approval(approved=True)` before connecting, so the approval step is bypassed automatically rather than being explicitly obtained.

Natural-Language Policy Violations

Low
Confidence
89% confidence
Finding
This markdown file is written entirely in Chinese and explicitly states that the unit is CNY rather than USD because 'the cost is RMB'. Under the policy rule for natural-language violations, this can be considered a locale/language constraint that is imposed without opt-in or an explicit justification that the skill is region-specific.

Natural-Language Policy Violations

Medium
Confidence
95% confidence
Finding
This markdown file uses Chinese throughout for headings, field descriptions, and operational notes, which effectively forces a specific language on readers. Under the policy rule, a language-specific constraint should either be optional for the user or explicitly documented as a justified regional limitation.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The code establishes a database connection and explicitly approves credential use, which is a sensitive operation involving access to protected data sources. Although there is an internal docstring comment, there is no user-facing prompt, log, or disclosure in the file warning that credentials will be used and a database connection will be opened.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.