ClawHub
Back to skill

Security audit

GEO品牌诊断

Security checks for vulnerabilities and agentic risk

Overview

This skill does what it advertises: it uses a logged-in browser session to query three AI platforms about a brand and generates local reports.

Install only if you are comfortable using authenticated Doubao, Yuanbao, and Tongyi sessions for automated brand-diagnosis queries. Use it on a trusted machine, avoid sensitive unpublished business information in inputs, and delete scripts/.chrome-profile if you want to clear saved login sessions.

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 (3)

Vague Triggers

Medium
Confidence
89% confidence
Finding
The trigger phrase set is broad and underspecified: generic phrases like “品牌诊断” or “诊断一下” can activate the skill outside the intended GEO-brand-analysis context. This can cause accidental invocation, unexpected collection workflows, or misuse in unrelated conversations, especially because the skill initiates browser-based platform interactions and reporting steps.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The script uses a persistent Playwright browser profile directory to retain login state and browsing/session artifacts locally, but it does not clearly disclose that these credentials and session data will be stored on disk. On shared workstations or poorly secured environments, another local user or process could reuse the stored sessions to access the third-party AI accounts.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The tool automatically sends brand names, locations, industries, and generated diagnostic queries to external AI platforms, but it does not provide an explicit privacy or data-sharing notice before transmission. If users input sensitive business information, that data may be exposed to third parties, logged by platform providers, or processed under external retention policies the user did not anticipate.

Static analysis

Detected: suspicious.dangerous_exec

Shell command execution detected (child_process).

Critical
Code
suspicious.dangerous_exec
Location
scripts/package.js:17