ClawHub
Back to skill

Security audit

Anti-Detection Writing

Security checks for vulnerabilities and agentic risk

Overview

This skill openly helps polish writing and optionally measure AI-detector scores, with safeguards against hidden posting, paid access, or unverifiable claims.

Install only if you are comfortable using a writing-polish skill that can also help measure and reduce AI-detector scores when you ask it to. Do not use it to violate site, school, workplace, or publication disclosure rules, and review any browser scans or public posting actions before they happen.

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 Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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
Findings (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
90% confidence
Finding
The skill references local files such as `references/measurement.md`, `references/evidence.md`, and `scripts/validate_receipt.py`, which implies file-read capability without declaring it. Undeclared file access weakens transparency and permission boundaries, making it easier for a skill to inspect local artifacts or adjacent files beyond what a user may reasonably expect.

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding
The skill is presented as a benign writing-polish tool, but the referenced behavior includes validating local GPTZero receipts, reading and hash-verifying local artifacts, checking UI/accessibility snapshots, and enforcing detector-score thresholds. That mismatch hides materially different operational behavior and increases the risk that users invoke verification, local artifact inspection, or anti-detection workflows without informed consent.

Static analysis

No suspicious patterns detected.