ClawHub
Back to skill

Security audit

Seedream 5.0 Lite 海报生成

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent AI poster-generation helper that uses a fixed AI Hive image model, with expected API-key setup, optional image upload, and local result saving.

Install only if you are comfortable giving the skill an AI Hive API key, uploading any reference images you choose to AI Hive/object storage, and saving generated outputs locally. Use only approved images and avoid pasting sensitive material into prompts or uploads.

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
91% confidence
Finding
The skill invokes shell commands, reads local files, writes output, accesses environment variables, and performs network requests, yet no permissions are declared. This creates a trust and review gap: operators and users cannot accurately assess the skill's capabilities, and a compromised or modified script could abuse those undeclared privileges to exfiltrate data or access unintended resources.

Description-Behavior Mismatch

Medium
Confidence
94% confidence
Finding
The skill claims to be a poster-generation skill, but it also exposes generic `task` and `upload` subcommands that allow users to query arbitrary generation tasks and upload media independently of the fixed poster-generation flow. This expands the skill’s operational scope beyond its stated purpose, increasing the risk of unintended data access, cross-task information exposure, and policy bypass if task IDs or media workflows are reused outside the intended model-specific path.

Static analysis

No suspicious patterns detected.