ClawHub
Back to skill

Security audit

Edit Video

Security checks for vulnerabilities and agentic risk

Overview

This skill provides disclosed video-editing guidance and does not show hidden execution, persistence, credential access, or destructive behavior.

Before installing, understand that this skill may guide both whole-frame restyles and some localized video edits. Users should confirm whether targeted actor, product, garment, or object swaps are intended to use this skill or a separate replace-in-video skill.

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 (3)

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest scopes this skill to whole-frame video restyling, but the body expands behavior to include localized/surgical edits and even routes users toward a distinct object-swap workflow. That creates authority drift: an orchestrator selecting this skill for global restyling could be induced to perform narrower edits outside the declared contract, causing misrouting, policy bypass, or execution of capabilities that should require a different reviewed skill.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill manifest says this capability is for whole-frame video restyling, but the examples teach users how to perform surgical element replacement and region-specific edits using a different model. That scope expansion can cause an orchestrating agent to misuse this skill for targeted modifications it was explicitly supposed to route elsewhere, undermining policy boundaries and tool-selection safety controls.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The documentation explicitly instructs routing one-region changes to Aleph inside the edit-video skill, directly contradicting the metadata that says element swaps should use replace-in-video. Contradictory routing guidance is dangerous because agents often prioritize detailed examples over high-level manifests, so this can systematically bypass intended separation of responsibilities between skills.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.