ClawHub
Back to skill

Security audit

Alibaba Cloud AI Video EMO

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward Alibaba Cloud EMO helper that writes local request artifacts and requires public media URLs, with no evidence of hidden persistence or malicious behavior.

Install only if you are comfortable using Alibaba Cloud Model Studio and providing public or provider-accessible URLs for portrait images and speech audio. Avoid sensitive biometric or voice media unless you control the hosting and understand the exposure. Keep the Alibaba API key in normal credential storage and review generated files under output/aliyun-emo if they may contain sensitive URLs or request metadata.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Rogue AgentSelf-Modification, Session Persistence
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
92% confidence
Finding
The skill performs file writes during validation and documents persistent output artifacts, but it does not declare any tool scope or permissions boundary. This creates an implicit capability gap where an agent may write files without explicit user-visible authorization, increasing the risk of unintended filesystem modification or misuse in broader agent environments.

Session Persistence

Medium
Category
Rogue Agent
Content
## Validation

```bash
mkdir -p output/aliyun-emo
python -m py_compile skills/ai/video/aliyun-emo/scripts/prepare_emo_request.py && echo "py_compile_ok" > output/aliyun-emo/validate.txt
```
Confidence
60% confidence
Finding
Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill requires input media to be hosted at public HTTP/HTTPS URLs, which means user images and audio must be exposed to third-party hosting or broadly accessible endpoints. Without a clear warning, users may unknowingly disclose sensitive biometric and voice data, creating privacy and data exposure risk that is heightened by the portrait-video use case.

Static analysis

No suspicious patterns detected.