ClawHub
Back to skill

Security audit

Review Workflow

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed code review and commit workflow that can inspect and modify a repository, but its behavior matches that purpose and shows no evidence of exfiltration or deception.

Install this only if you want an agent to run an opinionated full code review and commit-preparation workflow. Expect it to inspect current diffs and some untracked repo files, run local lint/test/audit tools, possibly auto-format files, and guide git operations; review changes before committing or pushing.

SkillSpector

By NVIDIA
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 (3)

Vague Triggers

Medium
Confidence
88% confidence
Finding
The trigger phrases are broad enough to fire on routine requests like 'review 一下' or '生成 commit', causing the skill to launch a multi-step workflow that reads repository state, invokes sub-skills, and runs shell commands. In an agent environment, overbroad activation increases the chance of unintended codebase inspection and command execution without sufficiently explicit user consent.

Vague Triggers

Medium
Confidence
93% confidence
Finding
The rule that the skill should proactively trigger even when the user only says '提交代码' or '走提交流程' is risky because it removes ambiguity handling and pushes the agent into autonomous review, scanning, and Git-adjacent actions. This makes unintended activation more likely and can expose sensitive untracked content or initiate impactful workflow steps the user did not explicitly request.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The skill advertises broad, everyday trigger phrases such as requests to commit code or prepare a submission, which can cause it to activate in situations where the user did not intend a high-impact Git automation flow. Because this skill can invoke Bash and perform staging and commits, over-broad invocation increases the chance of unintended repository changes and expands the attack surface for prompt-trigger abuse from adjacent conversation context.

VirusTotal

50/50 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.