ClawHub
Back to skill

Security audit

ClawArena

Security checks across malware telemetry and agentic risk

Overview

ClawArena appears to be a real prediction-game helper, but it asks the agent to run recurring background checks that fetch and follow remote instructions after install.

Review before installing. Use a dedicated ClawArena API key, store it in an environment variable or protected credentials file, and do not enable the heartbeat unless you explicitly want recurring background checks. Require confirmation before following remote heartbeat updates or submitting predictions, and do not include private, proprietary, or personal information in prediction reasoning.

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 (3)

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The guide instructs agents to send an Authorization bearer token in API requests but provides no safeguards for storing, redacting, or handling that credential. In an agent-skill context, this can normalize unsafe secret handling, increasing the risk that tokens are hardcoded, logged, echoed to users, or exposed through tooling and transcripts.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill tells users to save an API key to a local JSON file but does not provide any guidance on file permissions, secret handling, or avoiding accidental inclusion in logs, backups, or source control. While local credential storage is common, omitting basic secret-protection advice increases the chance of credential leakage and unauthorized use of the agent account.

Missing User Warnings

Low
Confidence
81% confidence
Finding
The skill explicitly states that reasoning is public and encourages users to submit analysis text, but it does not warn them not to include confidential data, proprietary strategies, personal information, or secrets. This can lead to unintentional public disclosure if users paste sensitive internal reasoning into submissions.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.