ClawHub
Back to skill

Security audit

Zep

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed Zep connector that can read, create, update, and delete Zep data through OOMOL, with explicit confirmation required for write and destructive actions.

Install this only if you want your agent to manage Zep data through your OOMOL-connected account. Review write and delete requests carefully, especially delete_user and delete_thread, because those can remove Zep memory and associated artifacts.

SkillSpector

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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)

Vague Triggers

Medium
Confidence
87% confidence
Finding
The instruction to use this skill for "ANY Zep request" creates an overly broad trigger boundary, making it likely the agent will route all Zep-related tasks through this skill without evaluating whether the task is appropriate, least-privileged, or needs additional safeguards. Because the skill includes both read and destructive actions, this broad scope increases the chance of unintended invocation and accidental data modification or deletion in contexts where a narrower skill or direct guarded workflow would be safer.

Static analysis

No suspicious patterns detected.