ClawHub
Back to skill

Security audit

Dialogflow CX to CX Agent Studio Migration Skill

Security checks across malware telemetry and agentic risk

Overview

This looks like a legitimate Google migration helper, but it deserves review because it uses broad Google credentials, can change cloud resources, and has unsafe endpoint/archive handling risks.

Install only if you are comfortable granting Google Cloud credentials that can export Dialogflow CX data and create or update CX Agent Studio resources. Use the default Google API endpoints, avoid custom base URLs unless fully trusted, run export-only first when possible, and protect or delete the generated export files because they may contain sensitive agent configuration.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
95% confidence
Finding
The skill declares and documents code-driven behavior that performs local file read/write and outbound network access, but it does not declare any corresponding permissions. Undeclared capabilities reduce reviewability and can mislead operators about what the skill can do, especially because it exports agent packages, writes artifacts to disk, and calls multiple Google APIs with broad cloud scopes.

Missing User Warnings

Medium
Confidence
82% confidence
Finding
The code calls zipfile.extractall() on exported archive contents without validating member paths. If a malicious or tampered archive is ever processed, this can allow path traversal ('zip slip') and overwrite arbitrary files under the executing user's permissions, which is more serious than merely lacking disclosure about local unpacking.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.