ClawHub
Back to skill

Security audit

software-spec-writing

Security checks for vulnerabilities and agentic risk

Overview

This is a documentation workflow skill that can strongly shape spec-first development, but its behavior is disclosed, file-scoped, and aligned with its stated purpose.

Install this only if you want agents to follow a strict spec-first workflow and maintain project specification files over time. Teams not working in Traditional Chinese should confirm they can review the generated documents or adapt the language expectations before relying on it.

SkillSpector

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (2)

Vague Triggers

Medium
Confidence
93% confidence
Finding
The skill declares very broad trigger keywords such as 規格, 文件, 需求, AC, coverage, and even general discussion contexts like making technical decisions or noticing external constraints. This can cause the skill to activate for many ordinary conversation and coding tasks, effectively hijacking task flow and forcing unnecessary spec-writing behavior before unrelated work. In an agent setting, overbroad activation is dangerous because it can change system behavior, increase prompt surface area, and create denial-of-task or misrouting conditions.

Natural-Language Policy Violations

Medium
Confidence
79% confidence
Finding
The skill metadata and content are written in Traditional Chinese and appear to impose that workflow without offering a user language choice or documenting a project-localization constraint. While this is not a code-execution risk, it can degrade safety and correctness by causing misunderstandings, reducing reviewability for teams using other languages, and making users accept outputs they cannot fully validate. The surrounding skill context makes this somewhat more concerning because it governs authoritative specification documents that downstream humans and agents may treat as source-of-truth.

Static analysis

No suspicious patterns detected.