Security audit

Evolve Loop Free

Security checks across malware telemetry and agentic risk

Overview

This is a local agent-memory skill, but it broadly activates and persistently records user corrections, preferences, and reflections across sessions.

Install only if you intentionally want a local, cross-session memory system. Review ~/evolve-loop regularly, avoid storing secrets or sensitive personal data, be careful with ZIP exports or git backups, and use the forget/delete workflow when memory should not persist.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (7)

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The trigger condition is so broad that the skill may activate in many unrelated conversations involving AI, LLMs, or agent workflows. In context, that matters because the skill performs persistent reads/writes and memory management, so overbroad activation can cause unintended retention or disclosure of user data without clear user intent.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The skill instructs the agent to create and persist local memory files immediately, but does not prominently warn users that their statements, corrections, and preferences may be retained across sessions. That creates a privacy risk because users may disclose personal or sensitive information under the assumption of ephemeral chat behavior.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The export/backup guidance encourages bundling all stored memory into a ZIP archive without a strong warning that the archive may contain accumulated personal, project, or sensitive operational data. Bulk packaging materially raises exposure risk because it consolidates all retained data into a portable artifact that is easier to exfiltrate or mishandle.

Ssd 3

Medium

Confidence: 92% confidence
Finding: The skill is explicitly designed to retain user-provided information across sessions and later surface it in future interactions. In this context, persistent cross-session recall increases the risk of privacy leakage, especially if multiple users, projects, or sensitive workflows share the same environment.

Ssd 3

Medium

Confidence: 93% confidence
Finding: The skill directs the agent to record broad classes of user statements such as habits, repeated reminders, and preferences into persistent memory. Because these categories can easily include personal or confidential information, automatic capture widens the chance of overcollection and later unintended disclosure.

Ssd 3

Medium

Confidence: 91% confidence
Finding: The transparency requirement tells the agent to cite prior memory contents and file locations in responses. While intended for traceability, this can expose stored user data or sensitive project structure in normal conversation, especially when the memory content itself contains confidential preferences, decisions, or prior corrections.

Ssd 3

High

Confidence: 97% confidence
Finding: The skill includes a command to package all memory files on request, enabling bulk disclosure of accumulated cross-session data. In a memory-retaining skill, this is particularly dangerous because a single export action can reveal the full corpus of stored user preferences, corrections, project history, and potentially sensitive operational context.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.