ClawHub
Back to skill

Security audit

ghost cms

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward Ghost CMS admin helper, but it should be used carefully because it can publish, update, delete, and upload content on a live site.

Install only if you want an agent to manage a Ghost site you control. Keep the config file private, rotate the Admin API key if it is exposed, confirm post IDs and titles before update/delete actions, prefer drafts unless publication is explicitly intended, and avoid passing untrusted or internal URLs as feature images.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Rogue AgentSelf-Modification, Session Persistence
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (6)

Context-Inappropriate Capability

Medium
Confidence
92% confidence
Finding
The documentation includes a realistic Ghost Admin API key in a shell export example, which normalizes embedding sensitive credentials directly in docs and configuration files. Even if the key is illustrative, it can be mistaken for a usable secret, copied into insecure storage, or encourage unsafe credential-handling practices for an API that can modify and delete site content.

Intent-Code Divergence

Medium
Confidence
91% confidence
Finding
The script advertises project isolation via a config-file-only workflow, but it still accepts user-controlled remote image URLs and fetches them with requests.get(). This creates an SSRF-style outbound request surface that can reach arbitrary hosts, including internal services, which undermines the claimed isolation model and may expose internal network resources or be abused for unexpected data egress.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill documents a delete capability for Ghost posts without any warning, confirmation step, or guidance to verify user intent before execution. In an agent setting, this increases the chance of accidental or unauthorized destructive actions that can remove published content and cause operational or reputational harm.

Missing User Warnings

Medium
Confidence
84% confidence
Finding
The document provides a delete command for posts without any warning that the action is destructive and may permanently remove content. In an agent skill context, omission of safety guidance increases the chance of accidental destructive operations against a live Ghost instance.

Missing User Warnings

Low
Confidence
76% confidence
Finding
The documentation shows direct publishing with `--status published` but does not warn that this immediately affects live site content. This can lead users or agents to unintentionally publish unreviewed material, causing reputational or operational harm on production sites.

Session Persistence

Medium
Category
Rogue Agent
Content
3. Click **"Add custom integration"**
4. Copy the **Admin API Key** (format: `id:secret`)

### 2. Create Configuration File

**唯一调用方式：自定义配置文件路径（项目隔离）**
Confidence
87% confidence
Finding
Create Configuration File **唯一调用方式：自定义配置文件路径（项目隔离）** Create a JSON config file for your site: Example: `/Users/ethan/.openclaw/workspace/projects/fuye/ghost-admin.config.json` ```json { "api_url"

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.