ClawHub
Back to skill

Security audit

错题整理巩固

Security checks for vulnerabilities and agentic risk

Overview

This skill analyzes student mistakes and can save them to a local mistake notebook, with the file-writing behavior clearly tied to that purpose.

Install this only if you want the assistant to keep a local study mistake notebook. Be aware that saved questions, answers, and analysis may include private schoolwork and will persist in the configured local directory until you delete them.

SkillSpector

Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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)

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
85% confidence
Finding

The skill explicitly instructs the agent to read and write local files via a notebook script and a workspace-backed mistake notebook, but no declared permissions are present. This creates a permission-model gap: the skill can influence persistent local data operations without transparent capability declaration, increasing the risk of unintended file access or writes if the skill is invoked on untrusted or ambiguous user input.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.