ClawHub

Humanpages

Security checks across malware telemetry and agentic risk

Overview

Humanpages appears to be a legitimate hiring integration, but it gives an agent access to sensitive human-profile data and real USDC payment workflows without enough built-in safety checks.

Install only if you want an agent to interact with a real human-hiring marketplace. Require manual approval before viewing full profiles, creating offers, sending or recording payments, changing payment streams, or posting reviews. Verify recipient identity, wallet address, network, amount, job status, and review text every time, and protect the Human Pages API key and webhook secret like account credentials.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Missing User Warnings

Medium
Confidence
86% confidence
Finding
The skill explicitly instructs the agent to save the returned API key and webhook secret, but provides no warning that these credentials are highly sensitive or guidance on secure storage and redaction. In an agent setting, this increases the chance the model echoes, logs, or persists secrets in unsafe places, which could enable unauthorized use of the Human Pages account and webhook spoofing.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The skill directs the agent to send USDC and manage payment streams, but does not prominently warn that blockchain payments are irreversible, may be sent to the wrong address/network, and can result in permanent financial loss. Because this skill is specifically for hiring and paying real humans, the operational context makes accidental or manipulated transfers materially more dangerous than a generic mention of payments.

VirusTotal

48/48 vendors flagged this skill as clean.

View on VirusTotal