ClawHub
Back to skill

Security audit

Project Watcher

Security checks for vulnerabilities and agentic risk

Overview

This skill mostly matches roadmap tracking, but its optional git hook can persistently send commit details to a hardcoded Feishu recipient and needs review before use.

Review and edit the Feishu recipient, credentials, and project configuration before installing. Only enable the post-commit hook in repositories where ongoing Feishu notifications are intended, and avoid using it where commit messages may contain sensitive customer, incident, ticket, or secret information.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Vague Triggers

Medium
Confidence
91% confidence
Finding
The trigger condition 'after exec / spawn subagent / tool call in the project directory' is extremely broad and can cause the skill to run after nearly any action. In an agent system, this can create excessive autonomous behavior, unintended file modifications such as roadmap churn, and prompt-surface amplification where ordinary operations repeatedly invoke nonessential workflow logic.

Missing User Warnings

Medium
Confidence
89% confidence
Finding
This post-commit hook automatically invokes a Python script with repository metadata including repo name, branch, commit hash, and commit message, and the script path suggests integration with an external notification system. Because git hooks run silently after local commits and there is no user-facing disclosure, consent, host validation, or scope restriction here, developers may unknowingly leak sensitive project metadata or secrets embedded in commit messages.

Static analysis

No suspicious patterns detected.