ClawHub
Back to skill

Security audit

Agent Memory Reflector

Security checks across malware telemetry and agentic risk

Overview

This is a local agent-memory logging and reflection tool; its main risk is that it saves prompts, responses, metadata, and reports as plaintext files on disk.

Install only if you are comfortable with interaction logs and reflection reports being saved in plaintext under .agent_memory. Avoid logging secrets, credentials, private conversations, or proprietary context unless you add redaction, restrictive file permissions, encryption, and a retention or deletion process.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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 (2)

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The tool persistently writes raw prompts, responses, and arbitrary metadata to `.agent_memory/memory.jsonl` without consent, minimization, encryption, or access controls. In an agent setting, these fields can easily contain API keys, credentials, personal data, system prompts, or other sensitive operational context, so local disk persistence materially increases confidentiality and compliance risk.

Missing User Warnings

Low
Confidence
88% confidence
Finding
The reflection report writes derived behavioral analysis to `reflections.jsonl`, which can expose sensitive operational patterns even if raw content is not directly included. These reports may reveal agent weaknesses, uncertainty trends, correction behavior, and usage patterns that could aid profiling or secondary disclosure if the filesystem is accessible.

VirusTotal

67/67 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.