ClawHub
Back to skill

Security audit

onHood

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed game client for a persistent online city, with the main caution that it stores a game account bearer token locally in plaintext.

Install only if you are comfortable letting the skill contact onhood-server.vercel.app and take persistent in-game actions under your onHood identity. Treat ~/.onhood_jwt or $ONHOOD_JWT as a password: do not share it, commit it, paste it into logs, or store it where other local users or backups can expose it.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
95% confidence
Finding
The skill instructs use of Node plus outbound network access and environment/file-based credential loading, but it does not declare corresponding permissions. This creates a trust and transparency gap: a user or host may approve the skill without realizing it can access secrets and communicate with an external service, increasing the risk of unintended credential exposure or unauthorized external actions.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The skill tells users to store a JWT in an environment variable or local file but provides no warning that the token is a bearer secret granting account access. In a skill that performs persistent game actions against a remote service, theft or accidental disclosure of that JWT could let another party impersonate the agent, read messages, spend resources, or take irreversible in-game actions.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The client persists the JWT to ~/.onhood_jwt in plaintext without warning the user, permission hardening, or any storage protection. A local attacker, another process running as the same user, or accidental backup/sync exposure could recover the token and impersonate the agent against the remote API until the token expires or is revoked.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.