ClawHub
Back to skill

Security audit

Contributor Guide Writer

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward contributor-guide generator that reads project files to produce a CONTRIBUTING.md and does not show hidden or high-risk behavior.

Before installing, understand that the skill may inspect normal repository files such as package manifests, CI config, and README content to make the guide accurate. That access is proportionate to its purpose, but use it in repositories where you are comfortable having the agent read project metadata and generate or update CONTRIBUTING.md.

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
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The example trigger phrase "Generate a contributor guide for this project" is broad enough that it can match common user requests in many repositories, causing the skill to activate frequently even when the user may only want general advice. Over-broad activation increases the chance of unintended repository inspection and automatic content generation based on workspace files, which can expose sensitive project context or cause the agent to take actions the user did not explicitly intend.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

The eval data explicitly expects a Chinese CONTRIBUTING.md for one prompt and an English CONTRIBUTING.md for another. This embeds language behavior in natural-language examples without any stated user opt-in or policy rationale, which can reflect a locale/language constraint in the skill's behavior.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.