Openclaw Admin

Security checks across malware telemetry and agentic risk

Overview

This is a coherent OpenClaw admin helper, but it gives broad gateway-management authority with weak guardrails around persistent routing, cron, and trigger changes.

Install only if you want an agent to help administer an OpenClaw gateway. Before allowing edits, require an explicit diff and approval for ROUTING.json, CRONS.json, TRIGGERS.json, agent prompt files, and openclaw.json, and keep a backup or rollback path. Inspect any local status.sh script and verify any npx restart command before execution.

SkillSpector

By NVIDIA

Vulnerability Patterns

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
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Missing User Warnings

Medium

Confidence: 87% confidence
Finding: The skill instructs the agent to execute `bash ./status.sh` for a full health check, but gives no visibility into what that script does or any warning that it may have side effects. In a security-sensitive admin skill, invoking an opaque shell script increases risk because a modified or malicious `status.sh` could run arbitrary commands, access secrets, or change system state under the guise of diagnostics.

VirusTotal

62/62 vendors flagged this skill as clean.

View on VirusTotal