ClawHub
Back to skill

Security audit

Linkedin Content Machine

Security checks for vulnerabilities and agentic risk

Overview

This is a LinkedIn content-generation package with marketing copy and samples, but no executable payload was present in the inspected artifact.

Before installing, treat this as a marketing/content-generation skill, not a fully verified working CLI: the package references missing command files. Do not provide private videos, confidential campaign plans, or sensitive channel strategy unless the publisher clarifies data handling and retention.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Credential Access

High
Category
Privilege Escalation
Content
The fix? A proper token lifecycle:

✅ Short-lived access tokens (15 mins)
✅ Refresh tokens with rotation
✅ Role-based claims for authorization
✅ Secure key storage (Azure Key Vault / AWS Secrets Manager)
Confidence
70% confidence
Finding
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Vague Triggers

Medium
Confidence
88% confidence
Finding
The README uses a very broad trigger phrase, Reply "yes", as the call to action. In agent or chat-integrated environments, generic affirmative phrases can be matched unintentionally during normal conversation, which may trigger actions or workflows without sufficiently informed user consent. The sales-oriented context increases risk because the phrase is positioned as an immediate activation step rather than a clearly scoped command.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The README asks users to provide a YouTube video link but gives no notice about how that data will be handled, whether it will be sent to external services, retained, or used for training or outreach. Even if the link is public, associated metadata, channel targeting, and user intent can constitute sensitive business information, and the absence of disclosure undermines informed consent and safe data handling.

Static analysis

No suspicious patterns detected.