ClawHub
Back to skill

Security audit

preview-outcome

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward guide for creating a temporary RooQuiz personality quiz preview, with the main caution that quiz content is uploaded to RooQuiz.

Install only if you are comfortable sending the quiz title, questions, answers, scoring mappings, and result text to RooQuiz to generate a temporary shareable preview. Do not use it with secrets, private business material, regulated personal data, or content you would not want hosted behind a preview link.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (3)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill explicitly directs users to send quiz content to a public third-party preview endpoint and emphasizes that no authentication is required, but it does not warn that quiz content is being transmitted off-platform and may include sensitive or proprietary material. This creates a real privacy and data-handling risk because users may assume the action is local or low-risk when it actually uploads content to an external service.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
95% confidence
Finding

The skill provides a concrete command to transmit raw JSON to an external domain, which is a genuine external data exfiltration path if users place sensitive prompts, internal content, or personal data into the quiz body. The surrounding context makes this more concerning because the endpoint is public, unauthenticated, and designed for easy browser sharing, increasing the chance that uploaded material becomes accessible beyond the original environment.

Content

Scanner excerpt · SKILL.md (reported line 45)May include surrounding context.

For example, with curl (write the JSON to a file first, or inline it with --data):

bash
curl -sS -X POST https://preview.rooquiz.com/api/preview-forms \
  -H 'Content-Type: application/json' \
  --data-binary @/tmp/preview-form.json

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
77% confidence
Finding

The documentation states language: "en_US" in the top-level example and presents it as the form language while noting the actual platform default is zh_CN. This can steer the agent toward forcing a specific locale without first checking the user's preference, and the file does not instruct the agent to ask or honor the user's desired language before choosing one.

Content

No source excerpt is available for this finding.