ClawHub
Back to skill

Security audit

Seedream 5.0 Lite 多参考图生成

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed image-generation helper that uploads user-chosen reference images to AI Hive, stores an API key locally, and downloads generated results.

Install only if you are comfortable sending the selected reference images, prompts, and generation parameters to AI Hive. Do not assume the source-lineage rules are technically enforced by the script; review prompts and input files yourself, especially for people, brands, logos, or licensed artwork. Protect the stored API key like any other credential.

SkillSpector

Vulnerability Patterns
  • 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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
88% confidence
Finding
The skill advertises operational capabilities that include shell execution, file access, environment access, network access, and local file writes, but it does not declare permissions or present a constrained security model. This increases the chance that users or hosting platforms will grant broader access than expected, especially since the skill also stores credentials locally and performs networked actions against an external API.

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding
The skill claims strict provenance/source-lineage controls and fixed model behavior, but the described behavior includes arbitrary parameter passthrough, standalone upload/task operations, browser-based key setup, and local credential storage without code-level enforcement of lineage rules. This mismatch can mislead users into trusting safety, authorization, and compliance guarantees that are not actually implemented, resulting in unauthorized uploads, weak auditability, or policy bypass.

Static analysis

No suspicious patterns detected.