ClawHub

TikTok Page

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed TikTok API helper that can read account data and publish videos, so it is appropriate only for users who intentionally want agent-assisted TikTok account management.

Install only if you want this agent to use TikTok OAuth credentials for the selected account. Grant the minimum scopes needed, protect ~/.config/tiktok-page/credentials.json with restrictive permissions, rotate tokens if exposed, and manually confirm the account, file path, caption, privacy level, and upload destination before any publish or upload action.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Intent-Code Divergence

Medium
Confidence
95% confidence
Finding
The documentation and agent rules claim that all outbound calls go only to open.tiktokapis.com, but the file-upload flow performs a PUT to a server-provided upload_url that may point to a different host. This mismatch can weaken trust boundaries and host allowlisting assumptions, causing the agent to send local file contents to an unvalidated destination if the URL is malicious, misissued, or unexpectedly broad.

Vague Triggers

Medium
Confidence
83% confidence
Finding
The skill is framed as a universal TikTok API skill that constructs calls inline based on user intent, which is broad enough to trigger on many generic TikTok-related requests. In an agent setting, vague invocation boundaries increase the chance of overbroad activation and unintended execution of sensitive actions such as reading account data or initiating publishing flows.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The skill supports account-changing actions like video publishing and local-file upload but does not present a prominent upfront warning or confirmation requirement. In an agent environment, this can lead to accidental posting or exfiltration of local media to external services based on ambiguous user requests.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal