ClawHub
Back to skill

Security audit

123

Security checks across malware telemetry and agentic risk

Overview

The skill appears to generate PowerPoint files as described, but its path handling can let caller-supplied inputs read templates or write output outside the skill folder.

Review before installing. Use only trusted XML and trusted PPT templates, and do not let untrusted users or automated prompts choose template_file or output_filename. The publisher should reject absolute paths and '..' traversal, constrain outputs to a dedicated folder, avoid overwriting existing files unexpectedly, and pin the dependency for reproducible installs.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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
91% confidence
Finding
The skill advertises file read/write capabilities through its behavior and documented inputs/outputs, but it does not declare any permissions. Undeclared filesystem access weakens sandboxing and reviewer visibility, and in this context could allow reading arbitrary local template files or writing generated PPTX output to unintended locations if inputs such as filenames are not tightly constrained.

Context-Inappropriate Capability

Medium
Confidence
97% confidence
Finding
The handler trusts caller-controlled `template_file` and `output_filename` and joins them directly with `skill_root` using `Path` semantics. An absolute path or traversal such as `../../...` can escape the skill directory, allowing unauthorized local file reads via `Presentation(str(template_path))` and arbitrary file writes via `prs.save(str(output_path))`, which is significantly more dangerous in an agent runtime that may have access to sensitive local files.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.