ClawHub
Back to skill

Security audit

财报深读

Security checks for vulnerabilities and agentic risk

Overview

The skill is a disclosed financial-research integration, but it asks the agent to fetch and run mutable remote code that can use a local Cue API key and consume credits.

Install only if you trust the Cue runner repository and are comfortable with the agent running code fetched from GitHub or Gitee, reading your local Cue API key, contacting cuecue.cn, and spending Cue credits after confirmation. Prefer reviewing or pinning the runner code before first use.

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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 (1)

Context-Inappropriate Capability

Medium
Confidence
93% confidence
Finding
The skill instructs the agent to clone or update a remote Git repository and then execute a Python runner from that repository. This expands the skill from financial analysis into arbitrary software retrieval and code execution, creating a supply-chain and remote-code-execution risk if the repository, mirror, or fetched content is compromised or changes unexpectedly.

Static analysis

No suspicious patterns detected.