ClawHub
Back to skill

Security audit

GitLab Meeting Assistant

Security checks for vulnerabilities and agentic risk

Overview

The skill creates sanitized GitLab issues only when explicitly requested, with clear project boundaries and confidentiality safeguards.

Select the intended GitLab project and use an integration that supports confidential issues. Explicit requests can write immediately without another confirmation. GitLab content is written in English, and sanitization depends on the agent following the instructions.

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
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • 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)

Authored GitLab content is forced into English without user opt-in.

Medium
Category
Output Handling
Confidence
99% confidence
Finding

The skill explicitly requires all authored GitLab content to be in English, even when the meeting uses another language. Allowing the requester to choose the language of spoken and chat replies does not provide a choice for GitLab content, and no justification for the English-only constraint is documented.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.