ClawHub
Back to skill

Security audit

huawei-cloud-mrs-clickhouse-sql-check

Security checks across malware telemetry and agentic risk

Overview

This is a local ClickHouse SQL checking skill with reliability and output-sensitivity caveats, but I found no hidden network access, credential use, persistence, or destructive behavior.

Install only if you are comfortable with a local checker that may echo full SQL text in output. Do not paste production secrets or credentials into SQL checked in shared terminals or CI logs, and do not rely on it as a complete validator for every ClickHouse statement type.

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 (3)

Description-Behavior Mismatch

Medium
Confidence
95% confidence
Finding
The registry only maps a limited subset of the declared StatementType values to concrete grammar definitions, despite the skill metadata and code comments claiming broad statement coverage. In a SQL validation skill, this can cause unsupported statements to be silently unvalidated, misclassified, or treated inconsistently, which may let unsafe or noncompliant SQL pass review and undermine downstream trust in the checker.

Intent-Code Divergence

Medium
Confidence
92% confidence
Finding
The module docstring states that it defines all statement types and their grammar structures, but many enum values have no corresponding grammar entry. This documentation/implementation mismatch is dangerous in a security-relevant validation component because operators or other code may rely on the claim of complete coverage and make security decisions based on incomplete analysis.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The CLI serializes and prints all token values, which includes the full SQL text and embedded literals such as credentials, personal data, or proprietary queries. In a review/checking tool context, users may pass production SQL on the command line or via automation, so echoing sensitive contents to stdout can leak data into terminals, logs, CI artifacts, or monitoring systems.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.dynamic_code_execution

Dynamic code execution detected.

Critical
Code
suspicious.dynamic_code_execution
Location
scripts/version_loader.py:50