ClawHub
Back to skill

Security audit

DFMA活动引导

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent DFMA meeting facilitator, but it can automatically monitor group chats, track participation, remind people, and publish a cloud report without clear opt-in or sharing controls.

Review before installing in real group chats. Use it only where the organizer and participants understand that the bot may read the DFMA discussion, track participation, send reminders, summarize content, and create a Feishu document. Confirm cloud document permissions and avoid uploading confidential design materials unless the workspace and link sharing are appropriately restricted.

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
  • 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)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The trigger logic is broad enough to activate on ordinary group-chat mentions of DFMA terms or any @ mention combined with a related request, which can cause the bot to begin monitoring conversation flow, tracking participation, and steering discussion without clear user intent. In a group setting, unintended activation can expose participant behavior and discussion content to automated collection and reporting, creating privacy and workflow risks.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill describes tracking participant activity, reminding non-responders, summarizing discussion content, and publishing a cloud document link, but it does not warn users or obtain consent before doing so. This creates a privacy and transparency issue because users may not realize their participation patterns and discussion content are being monitored, retained, and shared via a generated report.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
90% confidence
Finding

The entire skill file is written in Chinese and provides no indication that other languages are supported or that Chinese is required for a specific regional or organizational context. Under the policy rule, forcing a specific language without user opt-in can be a natural-language policy violation.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.