ClawHub
Back to skill

Security audit

Jev

Security checks for vulnerabilities and agentic risk

Overview

This skill is mostly a normal Jev connector wrapper, but it under-labels its only action as safe/read-like and includes fallback install commands that execute remote scripts directly.

Review before installing. Use it only when you intend to send text or JSON to Jev through OOMOL, avoid sensitive payloads unless that sharing is acceptable, and do not run the remote CLI installer commands without first verifying the installer source or using a safer installation method.

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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • 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
Findings (4)

External Script Fetching

High
Category
Supply Chain
Confidence
97% confidence
Finding

The skill instructs the agent to install software by piping a remotely fetched script directly into bash, which is a classic supply-chain and remote-code-execution risk. If the install endpoint, transport path, or hosting account is compromised, arbitrary code would execute immediately in the user's environment.

Content

Scanner excerpt · SKILL.md (reported line 55)May include surrounding context.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest and description frame the skill as suitable for 'searching and reading data,' but the only documented action is evaluate, which performs processing of user-supplied content rather than passive retrieval. This mismatch can mislead an agent into invoking the skill under a lower-risk assumption and bypassing appropriate review for non-read operations.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The instruction to use this skill for 'ANY Jev request' is overly broad and can trigger the skill for loosely related tasks without considering whether it is the least-privileged or most appropriate path. Broad routing language increases the chance of unnecessary external calls and accidental transmission of user data to the Jev connector.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The safety section states that untagged actions are reads, yet evaluate is untagged and is not a read/list/search operation. Incorrect safety labeling can cause downstream agents or users to treat execution as inherently safe, reducing scrutiny around what data is sent to the external service and what side effects or policy implications evaluation may have.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.