ClawHub
Back to skill

Security audit

Paper Doctor

Security checks for vulnerabilities and agentic risk

Overview

The skill appears aimed at academic paper review, but it grants shell execution without explaining why that power is needed or how it is controlled.

Review this skill before installing because it gives the agent shell-command authority that is not explained by the paper-review workflow. It would be safer if exec were removed or narrowly documented with explicit user consent for each command.

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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 (1)

Context-Inappropriate Capability

Medium
Confidence
93% confidence
Finding
The skill requests `exec` even though its stated purpose is reviewing and polishing academic papers, which can be fulfilled with file reading, editing, PDF extraction, and optional web search. Unnecessary shell execution expands the attack surface substantially: crafted paper content or prompt-injection embedded in documents could cause the agent to run local commands, exposing data or altering the environment.

Static analysis

No suspicious patterns detected.