ClawHub
Back to skill

Security audit

Taplio LinkedIn Post Writer

Security checks for vulnerabilities and agentic risk

Overview

The skill uses Taplio to research and draft LinkedIn posts, with user control over editing, scheduling, and immediate publication.

Connect only if you are comfortable sharing your topic and final draft with Taplio. Remove confidential details before using the skill. Review the final text and posting time carefully: scheduling leads to automatic publication, while immediate posting requires explicit confirmation.

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
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • 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
Findings (2)

The workflow sends the user's topic to Taplio without a user-facing privacy notice.

Medium
Category
Output Handling
Confidence
90% confidence
Finding

The skill automatically submits the user's topic to Taplio's inspiration search before drafting, without checking whether the raw idea contains confidential or unpublished information. The disclosed Taplio dependency makes this less suspicious than covert exfiltration, but connecting the service does not necessarily communicate that potentially sensitive brainstorming text will be transmitted as a search query before the user approves a post.

Content

No source excerpt is available for this finding.

The workflow extends draft creation into scheduling and potentially publishing LinkedIn posts.

Medium
Category
Excessive Agency
Confidence
94% confidence
Finding

The manifest describes generating variants and saving the chosen text as a draft. Line L83 additionally directs the agent to offer scheduling through schedule_draft and permits publish_draft with explicit confirmation, extending the workflow into public distribution. The confirmation requirement mitigates unauthorized publication but does not make these operations part of the declared draft-writing scope.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.