ClawHub
Back to skill

Security audit

Anthropic Usage

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed report-formatting helper with an optional local patch script; no hidden network, credential, or destructive behavior was found.

Install this only if you want help formatting or patching anthropic-report.py. Before applying the patch, run it with --dry-run, confirm the path is the intended report script, and keep the generated backup until you have tested the result.

SkillSpector

By NVIDIA
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
88% confidence
Finding
The skill exposes file read/write capabilities through its documented patch workflow but does not declare permissions, which can bypass reviewer and runtime expectations about what the skill is allowed to do. In an agent environment, hidden filesystem modification capability increases the risk of unauthorized local file changes, especially when the skill can patch arbitrary target paths.

Tp4

High
Category
MCP Tool Poisoning
Confidence
94% confidence
Finding
The stated purpose presents the skill as a report-format reference, but the implementation instructions include a patch tool that rewrites existing scripts on disk, creates backups, and applies code changes to arbitrary paths. This mismatch is dangerous because users or orchestrators may invoke the skill expecting passive formatting guidance while it actually enables active code modification, which can be abused for unintended or unauthorized changes.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal