ClawHub
Back to skill

Security audit

Beauty Karma

Security checks for vulnerabilities and agentic risk

Overview

This skill appears to do what it says: analyze a user-selected portrait with MiniMax and create local report files, with privacy cautions around photo upload and saved outputs.

Install only if you are comfortable sending selected portrait photos to MiniMax or a configured endpoint and saving generated report files that may contain the photo. Use mock mode for local demos, avoid untrusted image URLs or base URL overrides, and delete generated reports when they are no longer needed.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (5)

Lp3

Medium
Category
MCP Least Privilege
Confidence
88% confidence
Finding
The skill documentation indicates operational capabilities that require environment access and outbound network use, but it does not declare or surface those permissions explicitly. This can weaken user and platform transparency, making it easier for portrait images and API-backed processing to occur without clear permission scoping or review.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
This skill processes highly sensitive user-provided portrait images and the documentation does not clearly warn users that those images may be transmitted to a remote third-party model endpoint. Because facial images are sensitive personal data, lack of explicit disclosure and consent creates meaningful privacy and compliance risk.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill writes the generated analysis and report artifacts to disk, and those artifacts include the uploaded portrait image embedded as a data URL in SVG/HTML output. Because this is biometric-adjacent personal data plus derived profiling content, storing it locally without explicit disclosure, minimization, or retention controls creates a real privacy and data-exposure risk if the filesystem is shared, backed up, or later accessed by other users/processes.

Missing User Warnings

High
Confidence
98% confidence
Finding
The code transmits uploaded portrait images or image URLs to the external MiniMax API for analysis. Sending user photos to a third-party service without clear notice, consent, or data-handling controls is a genuine privacy vulnerability, especially because portrait photos can reveal highly sensitive personal characteristics even if the prompt asks the model not to infer them.

External Transmission

Medium
Category
Data Exfiltration
Content
- Model: `MiniMax-M3`.
- Real analysis requires `MINIMAX_API_KEY`.
- Default endpoint: `https://api.minimax.io/v1`.
- Override with `MINIMAX_BASE_URL` if needed.

## Safety Boundaries
Confidence
93% confidence
Finding
The skill is explicitly configured to send data to an external API endpoint, which is expected functionality, but it still creates a real data-exposure surface because user portrait images may leave the local environment. In this context, the risk is elevated because the transmitted content is biometric-adjacent and personally sensitive.

Static analysis

Detected: suspicious.env_credential_access

Environment variable access combined with network send.

Critical
Code
suspicious.env_credential_access
Location
scripts/run-skill.js:184