MemoryClaw Backup

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only helper for finding and using a separate MemoryClaw backup plugin, and its sensitive backup behavior is disclosed and aligned with that purpose.

Before installing, understand that this skill is only guidance; the separate MemoryClaw plugin will need access to your OpenClaw files and network service to perform backups. Verify that `clawhub:memoryclaw` is the intended plugin, review the plugin/source if you need stronger assurance, keep backup passphrases out of chat, and run login or restore steps directly in your terminal.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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 (2)

Description-Behavior Mismatch

Medium
Confidence
91% confidence
Finding
The README frames the repository as a discovery/onboarding skill while the published skill metadata presents it as a backup/restore assistant. This kind of security-relevant documentation mismatch can mislead users and downstream agents about what the skill actually does, causing inappropriate trust, incorrect invocation, or installation of a different component than expected.

Intent-Code Divergence

Medium
Confidence
88% confidence
Finding
The README says this repository is not the executable plugin and points users to a separate install target, while the skill description implies direct help with backup and restore operations. In a security-sensitive context involving backup and restore, this discrepancy can lead users or agents to rely on the wrong artifact, misunderstand the trust boundary, and accidentally install or authorize an unintended package.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal