Back to skill

Security audit

Memory.Local.Backup

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed local memory organizer, but users should treat its files as persistent private records.

Install only if you want a durable local memory folder. Review what gets written to ~/memory/, avoid storing secrets, credentials, financial, health, or highly sensitive personal details, and periodically delete or archive entries you no longer want retained.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (6)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The template explicitly encourages storing personal contact details, relationship context, communication preferences, and history in local markdown files, but provides no guidance on handling sensitive or personal data. This creates a privacy and data-protection risk because users may record personal information more broadly than intended and retain it indefinitely in unprotected files.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill asks whether built-in memory should be synced into a separate storage system, but it does not provide a meaningful privacy warning about copying previously shared personal information into a new, potentially more durable store. This is dangerous because users may not understand that prior preferences, contacts, or decisions could be duplicated into local files, expanding retention and exposure of sensitive data.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill directs the agent to copy prior memory and immediately store new user-provided information into a parallel long-term memory system without defining sensitivity boundaries, prohibited data classes, or consent requirements for high-risk information. In context, this memory skill is specifically designed for broad long-term retention, which makes overcollection and storage of personal, confidential, or regulated data more likely.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill explicitly instructs the agent to create and modify files under ~/memory/ as part of setup, but it does not require a clear user-facing disclosure or confirmation before filesystem writes occur. This is dangerous because it can cause unexpected persistence of user data on disk, create privacy surprises, and normalize silent local state changes by an agent.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
82% confidence
Finding

At L069 the document states 'Never modify built-in memory from this skill,' establishing separation between built-in memory and this skill. But L121-L125 instruct manual sync by reading from the agent's MEMORY.md, extracting sections, and writing them into ~/memory/sync/, which contradicts the earlier strong separation claim by treating built-in memory as a source for synchronization through this skill.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
81% confidence
Finding

The template instructs users to create persistent files under ~/memory but does not warn that data will be stored locally and remain on disk outside ephemeral agent context. While this is not inherently malicious, it can lead to unintended persistence of sensitive information and minor system hygiene issues if users assume the memory is transient.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.