Openclaw Echo Agent
Security checks across malware telemetry and agentic risk
Overview
This is a minimal echo/reference skill with no executable code or requested access, though it appears incomplete because it references a missing agent.py entry point.
This skill looks low-risk to install from the supplied artifacts because it contains only documentation and asks for no access. Before relying on it in workflows, verify whether the missing agent.py is intentional; as published, it may not actually run as an echo agent.
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
VirusTotal
63/63 vendors flagged this skill as clean.