ClawHub
Back to skill

Security audit

课程字幕校对

Security checks for vulnerabilities and agentic risk

Overview

The subtitle workflow and setup are clearly disclosed, with user approval required for paid processing and subtitle changes.

Consider whether your course materials may be uploaded to this provider and retained for 30 days. Setup stores a platform key and restarts the gateway; use the documented product-specific configuration where appropriate. Confirm current prices before paid processing. All five bundled files were reviewed, but the linked live documentation could not be independently retrieved.

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
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The supplied commands perform platform configuration and service administration, not transcription, subtitle proofreading, human review, or subtitle export. The description does not disclose these administrative changes, and the chunk provides no connection establishing them as necessary supporting steps for the declared workflow. There is no evidence of credential exfiltration in the supplied code.

Content

No source excerpt is available for this finding.