ClawHub
Back to skill

Security audit

Tvs Inksnow Arch

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed architecture-planning helper that writes project architecture docs and Cursor rules only after user approval.

Install only if you want the agent to help formalize architecture decisions and create persistent Cursor rules for future coding behavior. Review the generated RFC and .cursor/rules files before approving them, because alwaysApply rules can steer later agent edits.

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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
95% confidence
Finding

The skill instructs the agent to probe for the presence of another skill in user-level and project-level paths, which reveals details about the local skill installation and agent environment. Even though this is framed as feature detection, it performs environment enumeration on paths outside the immediate project and can leak filesystem layout or installed capabilities to downstream logic or user-visible output.

Content

Scanner excerpt · SKILL.md (reported line 42)May include surrounding context.

md
按以下顺序检查：

1. `Glob ~/.cursor/skills/tvs-deep-interview/SKILL.md`（用户级 skill 仓库）
2. `Glob .cursor/skills/tvs-deep-interview/SKILL.md`（项目级 skill）
3. 当前 Agent 环境的可用 skill 列表（如有元接口可查）

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
95% confidence
Finding

This instruction continues the same enumeration behavior by probing a project-local skill path and then consulting the agent's available skill list. Such discovery logic expands the agent's knowledge of installed tools and repository structure beyond what is needed for architecture guidance, creating unnecessary information disclosure and a reconnaissance primitive for chaining with other attacks.

Content

Scanner excerpt · SKILL.md (reported line 43)May include surrounding context.

md
按以下顺序检查：

1. `Glob ~/.cursor/skills/tvs-deep-interview/SKILL.md`（用户级 skill 仓库）
2. `Glob .cursor/skills/tvs-deep-interview/SKILL.md`（项目级 skill）
3. 当前 Agent 环境的可用 skill 列表（如有元接口可查）

### 情况 A：已安装