Security audit

memory

Security checks across malware telemetry and agentic risk

Overview

This is a coherent memory skill, but it broadly tells the agent to record and retain user, project, and people details without clear privacy limits or user control.

Review this skill before installing if you work with confidential, personal, legal, medical, financial, or credential-related information. It does not appear to exfiltrate data or run code, but it is designed to create long-lived memory files and archives, so users should add or require rules for consent, redaction, sensitive-data exclusions, and retention limits.

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 (4)

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: The instruction to 'write everything down' establishes a broad mandate to persist potentially sensitive user, project, or personal information without any data minimization, consent, or secrecy boundaries. In a memory skill, this context makes the issue more dangerous because the skill is explicitly designed to create durable cross-session records, increasing the chance of retaining secrets, personal data, or regulated information.

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: The skill directs automatic creation, updating, pruning, and archiving of workspace files, but does not require user awareness or confirmation before modifying persistent data. In this skill's context, these are core operational behaviors, so silent file mutations could cause unintended data retention, deletion of useful context from indexes, or unsafe archival of sensitive content.

Ssd 3

Medium

Confidence: 97% confidence
Finding: A broad instruction to persist all meaningful conversation content into memory files creates a strong risk of over-collection and long-term storage of sensitive information. Because the skill is specifically a structured memory system with daily logs and promotion rules, it operationalizes persistent retention rather than incidental note-taking, which increases exposure if secrets or personal data are discussed.

Ssd 3

Medium

Confidence: 95% confidence
Finding: The rule 'If it matters enough to say, it matters enough to log' encourages indiscriminate persistence of anything deemed important, which can easily include confidential or personal data. In a memory-management skill, this is especially risky because the rule acts as a default behavioral policy across sessions and can normalize storing sensitive content without user consent.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.