ClawHub
Back to skill

Security audit

Dataify Google Maps

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed, read-only Google Maps search wrapper for Dataify, with minor documentation inconsistencies but no evidence of hidden or destructive behavior.

Install only if you are comfortable with your Google Maps search terms, place identifiers, and any location anchors being sent to Dataify using your DATAIFY_API_TOKEN. Avoid using it for sensitive personal locations unless you intentionally want that query submitted to the external API.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Lp3

Medium
Category
MCP Least Privilege
Confidence
95% confidence
Finding
The skill instructs the agent to read environment state, invoke a local Python script, and make outbound network requests, yet no permissions are declared. This creates a transparency and policy-enforcement gap: users and hosting systems may not realize the skill can access tokens and call external services, increasing the risk of unintended data exposure or execution in environments that would otherwise restrict it.

Intent-Code Divergence

Medium
Confidence
84% confidence
Finding
The workflow first requires a confirmation step before calling the API, but the later policy says clear low-risk requests should execute immediately without confirmation. Contradictory control-flow instructions can cause an agent to behave unpredictably, sometimes skipping intended review or sometimes exposing extra internal details while seeking confirmation.

Intent-Code Divergence

Medium
Confidence
90% confidence
Finding
The mandated confirmation table reveals fixed/internal/default parameters such as engine selection, response-format flags, offsets, and other implementation details, despite later instructions saying not to expose them. Exposing these internals can confuse users, leak unnecessary operational detail, and make prompt steering easier by teaching an attacker the exact parameter surface the agent will honor.

Missing User Warnings

Medium
Confidence
87% confidence
Finding
The skill authorizes direct execution of user requests against an external API without clearly surfacing that the user's search terms and location-related inputs will be transmitted to a third party. This can cause unintended disclosure of potentially sensitive queries or geographic context, especially because the surrounding policy encourages automatic execution for low-risk requests.

Static analysis

No suspicious patterns detected.