ClawHub
Back to skill

Security audit

Mempalace Bwm

Security checks across malware telemetry and agentic risk

Overview

This local memory skill is coherent, but it automatically preserves conversation-derived memories across sessions without clear consent or retention safeguards.

Install only if you intentionally want persistent local memory. Before use, set explicit rules for what the agent may save, avoid storing secrets or sensitive personal data, and require confirmation before diary writes, knowledge-graph changes, or drawer deletion.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The skill explicitly instructs the agent to write a diary entry after every session, recording what happened and what was learned, but it does not require user awareness or consent before persisting conversation content. Even though storage is local, this can silently retain sensitive personal, project, or credential-adjacent information across sessions, creating a privacy and data-retention risk.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The documented `mempalace_delete_drawer` capability enables deletion of stored memory content, but the skill gives no warning to seek confirmation or explain consequences before destructive actions. An agent following this skill could delete user data accidentally, based on ambiguous instructions, or through prompt manipulation routed into tool usage.

Ssd 3

Medium
Confidence
97% confidence
Finding
The protocol directs the agent to persist detailed session summaries and learned information from every conversation into long-term local storage. In context, this makes the skill more dangerous because it is framed as mandatory behavior every session, increasing the chance that sensitive or unnecessary data is routinely retained without minimization or consent.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.