hume-x

• Outcome: Review
• Latest audit: May 29, 2026 at 9:22 PM UTC
• Version:
  1.0.0

Overview

This skill is a coherent X/Twitter automation wrapper, but it asks agents to use browser session cookies and can post, delete, and change account relationships without clear guardrails.

Install only if you trust the external hume-x npm package and are comfortable giving an agent access to your X/Twitter session. Prefer explicit credentials over browser-cookie discovery, consider a dedicated account, and require manual confirmation before posting, deleting, following, unfollowing, muting, blocking, pinning, retweeting, or scheduling tweets.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (3)

Missing User Warnings

Medium

• Confidence: 90% confidence
• Finding: The skill explicitly documents multiple credential sources, including direct access to browser cookies from Firefox or Chrome, but provides no warning that these tokens are highly sensitive session credentials. In this context, cookie-based auth can let the tool act as the user on X/Twitter, so normalizing browser-cookie access without clear consent and handling guidance increases the risk of credential misuse or unintended account compromise.

Missing User Warnings

Medium

• Confidence: 84% confidence
• Finding: The skill advertises posting, replying, quoting, and deleting tweets without any warning that these are account-impacting actions performed as the authenticated user. In an agent setting, this creates a meaningful risk of unintended public posts or destructive actions if invoked without strong user confirmation.

Missing User Warnings

Medium

• Confidence: 85% confidence
• Finding: The skill includes follow, unfollow, mute, and block operations without explaining that these alter the user's social graph and moderation state. In an automated-agent context, omission of such warnings can lead to unintended reputation, relationship, or account-management consequences.

VirusTotal

63/63 vendors flagged this skill as clean.