ClawHub
Back to skill

Security audit

OpenClaw Agent Control Skill

Security checks across malware telemetry and agentic risk

Overview

This deployment skill is purpose-aligned, but it runs unpinned remote code and starts persistent network services with limited warning or cleanup guidance.

Install only if you trust the GitHub repository and intend to run a local service deployment. Before running it, review or pin the source revision, set PROJECT_DIR to a non-root path when possible, bind the backend to localhost unless external access is needed, and know how you will stop the backend/frontend processes afterward.

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
89% confidence
Finding
The skill explicitly instructs the user to run a deployment script that clones a remote repository into /root and starts backend/frontend services, but it does not warn that this will modify system state, execute code from an external source, and run processes on the local host. In an agent-skill context, this is risky because users may treat the skill as a safe helper and trigger privileged changes without understanding the trust and persistence implications.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The usage instructions tell the user to execute a deployment script directly (`bash scripts/deploy_project.sh`) but provide no warning about what system changes it will make, what privileges it may require, or what external resources it may access. In a skill intended to deploy backend and frontend components, this increases the risk of users running potentially destructive or network-retrieving actions without informed consent or prior review.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The script automatically starts a long-lived backend with nohup and then deploys/restarts the frontend without any confirmation, teardown logic, or explicit warning that it is making persistent system changes. In a skill context, this is security-relevant because running the skill causes services to remain active and exposed after the session ends, increasing the chance of unintended resource use or network exposure.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.