ClawHub
Back to skill

Security audit

PinePaper

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly a coherent PinePaper canvas reference, but it exposes destructive canvas actions and an undefined training-data export tool without clear safeguards or scope.

Review before installing if you will use PinePaper with sensitive or valuable canvas content. Agents should confirm before clearing, resetting, or deleting selected items, and should not use the training-data export tool unless the user explicitly asks and understands what will be exported.

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
  • 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 (3)

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The skill describes destructive operations such as `agent.reset`, `app.clearCanvas()`, and undo-history-affecting actions without warning that they can erase existing work. In an agent context, that omission increases the chance an LLM invokes these commands during normal task flow, causing silent data loss or disruption of the user's project.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
Listing `pinepaper_delete_selected` among MCP tools without any safety guidance exposes a direct deletion capability that an agent may call on the active selection with no user awareness. Because selections can be broad or accidental, this creates a realistic path to destructive actions and permanent loss if undo is unavailable or not used promptly.

Context-Inappropriate Capability

Medium
Confidence
94% confidence
Finding
The skill advertises a `pinepaper_export_training_data` tool even though the surrounding document is framed as a graphics/canvas creation reference, not a data collection or model-training workflow. Exposing training-data export without clear scope, consent, or data-boundary documentation can enable unintended extraction of user-created content and metadata for secondary use.

Static analysis

No suspicious patterns detected.