ClawHub
Back to skill

Security audit

the skill to browse X

Security checks for vulnerabilities and agentic risk

Overview

This skill is openly for X/Twitter browsing and account actions, but it gives agents one-click instructions for visible account actions without requiring confirmation for all of them.

Install only if you want an agent to operate X/Twitter through your browser session. Treat it as write-capable: ask for read-only browsing unless you explicitly want account actions, and require a final confirmation before any post, like, repost, reply, or follow.

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 (2)

Vague Triggers

Medium
Confidence
89% confidence
Finding
The skill description is broad enough to activate for many generic X/Twitter-related requests, including state-changing actions such as posting, liking, reposting, replying, and following. In an agent environment, vague activation criteria can cause the skill to be selected too readily, increasing the chance of unintended account actions or privacy-sensitive browsing without explicit user intent.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill includes multiple write actions—like, repost, reply, and follow—but only explicitly requires confirmation before posting. Without equivalent confirmation and warning guidance for these other actions, an agent could perform irreversible or user-visible account actions based on ambiguous prompts, causing reputation damage, accidental engagement, or unauthorized account activity.

Static analysis

No suspicious patterns detected.