ClawHub
Back to skill

Security audit

esa

Security checks for vulnerabilities and agentic risk

Overview

This skill coherently lets an agent operate esa through the oo CLI, with disclosed write/delete actions and user confirmation requirements.

Install only if you intend to let the agent work with your esa teams through OOMOL. Review exact payloads before approving write or destructive actions, especially delete_comment and rollback_post_revision, and verify the oo CLI installer source before first-time setup.

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)

Intent-Code Divergence

Medium
Confidence
92% confidence
Finding
Several actions that appear read-only (`get_comment`, `get_post`, `get_post_backlinks`, `get_post_comments`, `get_post_summary_prompt`) are labeled `[write]`, while the Safety section states that only tagged actions change state. This inconsistency can mislead an agent or operator about which commands require confirmation, causing unnecessary approval prompts at best and unsafe trust in the tagging model at worst if future state-changing actions are mislabeled the other way.

Static analysis

No suspicious patterns detected.