safeclaw

Security checks across malware telemetry and agentic risk

Overview

This appears to be a local OpenClaw administration and diagnostics skill whose sensitive config access is disclosed and mostly bounded by safety rules.

Install only if you want an agent to help administer an OpenClaw installation. Expect it to run local diagnostic commands and inspect OpenClaw-related config/log paths, but do not let it read secrets files or session contents unless you explicitly approve a narrow need; review any proposed config or service changes before confirming.

SkillSpector

By NVIDIA

Vulnerability Patterns

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

Findings (1)

Vague Triggers

Medium

Confidence: 89% confidence
Finding: The skill is effectively invocable for any user request to 'run security checks on their configuration' without constraining what qualifies as a valid config path or repository scope. In an agent setting, this broad trigger plus direct shell command construction can cause the skill to run on unintended files or ambiguous user input, increasing the chance of unauthorized file inspection or unsafe command invocation paths.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal