ClawHub
Back to skill

Security audit

code-review

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent PR-review helper, but it can change GitHub PR state from a broad natural-language trigger without confirmation for every external action.

Install only if you want ordinary PR-review requests to route into this multi-agent workflow. Before use, consider editing the trigger to require `/code-review` or explicit multi-agent review, and require confirmation before adding Copilot as a reviewer or posting GitHub comments.

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
  • 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 (1)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill explicitly states that generic natural-language phrasing like "review this PR" is the trigger, which is broad enough to match ordinary user requests and cause this orchestrator to run unexpectedly. Because the skill can initiate reviewer assignment and guide follow-on actions like commenting, pushing, and issue creation, accidental invocation expands the chance of unintended external actions or unnecessary tool use.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.