Security audit

Agent Memory Hardened

Security checks for vulnerabilities and agentic risk

Overview

This is a local persistent-memory skill with real privacy considerations, but the reviewed artifacts are coherent and do not show hidden exfiltration or unsafe automatic behavior.

Before installing, verify the package name matches the reviewed skill, treat the local memory database as sensitive, avoid storing secrets or private personal data without consent, use expirations or cleanup for stale facts, and review recalled memory before letting an agent take important actions.

SkillSpector

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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

Findings (4)

Missing User Warnings

Medium

Confidence: 85% confidence
Finding: The README explicitly promotes persistent storage of facts, lessons, and entity data about people across sessions, including examples like boss preferences and personal attributes, but it does not warn about privacy, consent, retention, or handling of sensitive data. In an agent-memory skill, this omission is meaningful because users may deploy it to store personal or organizational information indefinitely without safeguards, increasing the risk of privacy violations, over-retention, and accidental misuse.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The skill is described in very broad terms as a persistent memory system for facts, lessons, and entities across sessions, with no strong trigger boundaries or activation constraints. In an agent environment, that can cause the skill to be invoked for many common interactions and silently accumulate cross-session data, increasing the chance of privacy violations, memory poisoning, and inappropriate retention even though some guardrails are present.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: This skill persistently stores arbitrary facts, lessons, and entity attributes to a local SQLite database under the user's home directory by default, but provides no built-in notice, consent flow, or retention disclosure. In an agent context, stored content can easily include sensitive personal, business, or behavioral data, so silent persistence increases privacy and compliance risk even if there is no direct remote exploit.

Missing User Warnings

Medium

Confidence: 84% confidence
Finding: The skill exposes permanent deletion operations such as forget() and forget_stale() that remove records without confirmation, soft-delete protection, or recovery support. In an agent memory system, mistaken invocation or prompt-driven misuse could irreversibly erase history, degrading integrity, auditability, and reliability of agent behavior.

Static analysis

No suspicious patterns detected.