🧠 Memory Never Forget 🧠

• Outcome: Pass
• Latest audit: May 29, 2026, 9:34 PM
• Version:
  4.1.3

Overview

This is a disclosed local memory skill that stores and recalls user/project context, with privacy considerations but no evidence of hidden execution, exfiltration, or malware.

Install only if you want OpenClaw to keep local memories about you, your preferences, and your projects across sessions. Periodically inspect and prune MEMORY.md, USER.md, memory/, knowledge/, and todos.md, especially for confidential, sensitive, or outdated information.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (5)

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The skill explicitly instructs proactive persistence of user profile, feedback, project decisions, and external references to disk, but it does not present a clear consent, minimization, retention, or privacy notice to the user. This creates a real privacy/security issue because sensitive personal or project data may be stored long-term without the user's informed awareness, increasing the risk of over-collection, unintended retention, and later disclosure through memory recall or file access.

Missing User Warnings

Medium

• Confidence: 91% confidence
• Finding: This design explicitly directs storage of personal details, commitments, and project context in persistent files, but the document does not require clear user notice, consent, data minimization, or retention controls for sensitive profile data. In an agent memory skill, this is risky because it can normalize silent collection of personal information and long-lived conversation history beyond what users reasonably expect.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The documented USER.md and daily log files describe persistent storage of identity, preferences, timezone, and detailed conversation logs, yet there is no parallel requirement to disclose that persistence to the user or obtain approval. In this context, the skill is specifically a memory system, which makes the omission more dangerous because routine operation would encourage systematic accumulation of user data across sessions.

Missing User Warnings

Medium

• Confidence: 87% confidence
• Finding: The templates instruct the agent to record user information, commitments, corrections, and knowledge into persistent files, but they provide no guidance on consent, data minimization, sensitivity handling, or retention limits. In a memory skill, this omission can normalize storing personal or behavioral data by default, creating privacy risk and making downstream misuse or over-collection more likely.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The User Profile Update template explicitly structures storage of user-identifying and preference data in USER.md without any warning, approval step, or restriction on what fields may be saved. Because this skill is specifically a long-term memory system, the template increases the chance of persistent collection of personal data beyond user expectations, which can expose sensitive information or enable profiling.

VirusTotal

64/64 vendors flagged this skill as clean.