ClawHub
Back to skill

Security audit

CSDN 文章发布

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward CSDN publishing helper, but it can post publicly under the user's logged-in account without a final confirmation checkpoint.

Install only if you are comfortable with the agent using your logged-in CSDN browser session to publish posts. Before using it, ask the agent to show the final title and content and wait for your explicit confirmation before clicking publish.

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 (2)

Tp2

High
Category
MCP Tool Poisoning
Confidence
85% confidence
Finding
Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill automates an externally visible publish action to a third-party platform without requiring an explicit final user confirmation immediately before clicking the publish button. This creates a risk of unintended or premature publication of sensitive, inaccurate, draft, or policy-violating content under the user's authenticated account.

Static analysis

No suspicious patterns detected.