ClawHub
Back to skill

Security audit

Taplio LinkedIn Engagement Tracker

Security checks for vulnerabilities and agentic risk

Overview

The skill reviews LinkedIn engagement through Taplio and permits optional draft changes, with explicit approval required before scheduling comments.

Requires a connected Taplio account and access to your comment content and profile analytics. Choose to leave drafts unchanged for reporting only. Optional rewrites modify saved drafts; approving a commit schedules a public comment.

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
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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)

The audit skill also modifies and schedules LinkedIn comments.

Medium
Category
Excessive Agency
Confidence
97% confidence
Finding

The manifest at L03 describes reading comment history and analytics to return a status board, failure fixes, and recommendations. L85 instead instructs the agent to call update_comment_draft to rewrite remote drafts and commit_comment_draft to schedule comments for posting. Requiring the user's explicit go for commits limits unintended posting, but these write operations still exceed the advertised audit-and-report purpose.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.