ClawHub
Back to skill

Security audit

Discord社区免费

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly a read-only Discord OAuth helper, but it needs Review because it combines personal Discord data access with an under-scoped callback URL and broad local tool declarations.

Install only if you are comfortable granting a skill access to Discord OAuth profile and guild information. Avoid using callback_url unless you control the destination and understand what data will be sent, and consider limiting the agent's local write/exec authority because the skill does not explain why those tools are needed.

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Intent-Code Divergence

Medium
Confidence
94% confidence
Finding
The skill repeatedly claims all operations are "safe" and side-effect-free, yet it accepts a callback_url for asynchronous completion. A callback introduces outbound network behavior and can transmit execution status or user-derived data to an arbitrary destination, which contradicts the read-only/safe framing and can mislead users and agents into permitting unexpected exfiltration paths.

Context-Inappropriate Capability

Medium
Confidence
92% confidence
Finding
For a Discord read-only query assistant, callback URL support adds an unnecessary outbound communication channel unrelated to core functionality. An attacker or careless integrator could direct results or metadata to an external endpoint, enabling data leakage, SSRF-like behavior, or covert signaling under the guise of a harmless lookup skill.

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The skill describes access to sensitive OAuth-backed data such as user profile, email, locale, guild membership, and OIDC claims, but provides no privacy warning or data-handling guidance. This can cause users or downstream agents to collect, expose, log, or retain personal data without informed consent or appropriate safeguards.

Static analysis

No suspicious patterns detected.