ClawHub
Back to skill

Security audit

Craft

Security checks for vulnerabilities and agentic risk

Overview

This Craft skill is not malicious, but it gives agents broad production write/delete guidance and includes a risky delete-and-recreate rewrite path users should review carefully.

Install only if you are comfortable letting the agent modify real Craft data. Before using it on existing documents, require explicit confirmation for deletes, full rewrites, collection item changes, task deletion, or batch writes, and keep a recoverable copy of important content.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (3)

Intent-Code Divergence

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill first emphasizes that it operates on production data and marks permanent deletions as unsafe, then later recommends deleting and recreating documents as the 'fast' rewrite path. That contradiction can normalize destructive behavior and lead an agent to irreversibly remove user content under the guise of routine editing, especially because the server is explicitly connected to real production data.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest and description frame the skill as document creation/formatting, but the body grants much broader destructive and administrative capabilities, including deletion, collections, whiteboards, comments, and other mutations. This mismatch can cause users or orchestrators to invoke the skill with a narrower trust model than the skill actually requires, increasing the chance of unintended high-privilege operations on production data.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger description is extremely broad, covering essentially any Craft document or note creation, editing, or formatting scenario. Broad triggers increase the likelihood of unintended invocation in contexts where the agent only needed read-only help or minor formatting, but the skill also exposes write, delete, and other mutable operations against production data.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.