Security audit

Openclaw Automation Guide

Security checks across malware telemetry and agentic risk

Overview

This is a documentation-only automation guide, but copied examples could automate file access or outbound messages if users enable them.

Before installing or copying these examples, treat them as templates rather than safe defaults. Use dedicated non-sensitive folders, allowlist destinations, review payloads before publishing or forwarding, test automations manually, and disable scheduled jobs you no longer need.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Missing User Warnings

Medium

Confidence: 87% confidence
Finding: The skill provides concrete automation examples for publishing content, forwarding messages across platforms, and sending email without any warning, consent boundary, or guidance on what data may be transmitted externally. In an automation-focused skill, users may enable these workflows as-is, creating a realistic risk of unintended data disclosure, spam, or unauthorized external actions.

Missing User Warnings

Medium

Confidence: 90% confidence
Finding: The file-trigger example watches a local Documents folder and processes matching PDF files without disclosing that the automation may access personal files or move their contents into downstream workflows. Because this skill is specifically about unattended automation, the omission makes accidental exposure or processing of sensitive local documents more likely.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.