ClawHub
Back to skill

Security audit

Eu Ai Compliance

Security checks for vulnerabilities and agentic risk

Overview

This skill is a simple EU AI Act compliance reference that points users to external web endpoints, with privacy cautions users should consider before sending system details.

Before using the endpoint examples, avoid sending secrets, personal data, regulated data, proprietary controls, or detailed internal architecture unless you have approval and understand the provider's handling of submitted data. Treat the output as compliance triage and documentation assistance, not legal advice.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The documented API usage encourages sending AI-system descriptions and agent identifiers to external domains but does not warn that this transmits potentially sensitive operational, business, or regulated information off-platform. Because the skill is specifically about compliance and technical documentation, users are more likely to provide detailed internal data, increasing the risk of confidential information leakage and unintended third-party processing.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The 'when to use this skill' section is broad and lacks boundaries about what data is appropriate to send, what systems are in scope, and when the external service should not be used. In a compliance context, users may submit sensitive AI-system descriptions, internal controls, or identifiers to a third-party endpoint without realizing the disclosure or validating whether the tool is suitable for regulated or confidential environments.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.