ClawHub
Back to skill

Security audit

PRD Creator

Security checks for vulnerabilities and agentic risk

Overview

This skill is a documentation helper that creates PRD and task Markdown files, with no evidence of hidden code execution, credential use, or data exfiltration.

Before installing, expect the skill to ask clarifying questions and create PRD/task Markdown files under `.project/prd/`. Review those files before handing them to a coding agent, and add `.project/` to `.gitignore` if the documents may contain private product, customer, or business details.

SkillSpector

Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
87% confidence
Finding
The trigger description is unusually broad and overlaps with many common software-documentation requests (e.g. feature spec, implementation plan, project brief), which can cause the skill to activate outside its intended narrow use case. Overbroad routing can redirect unrelated user requests into a heavyweight workflow, increasing the chance of incorrect tool use, unintended file creation, and policy- or instruction-bypassing behavior through misclassification of user intent.

Natural-Language Policy Violations

Medium
Confidence
83% confidence
Finding
Mandating English-only output regardless of the user's language can override user intent and create unsafe mismatches in comprehension, review, and approval workflows. In a requirements-document skill, language rigidity can lead users to accept documents they cannot fully validate, which increases the risk of misunderstandings being handed off directly to coding agents for implementation.

Static analysis

No suspicious patterns detected.