Security audit

Mneme

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent persistent-memory skill, but it gives the agent broad discretion to save long-term plaintext memories without clear consent or sensitivity limits.

Install only if you want project memories stored in local .mneme files. Do not let it store secrets, tokens, passwords, regulated data, or unnecessary personal information; review .mneme files periodically and delete or supersede entries you no longer want retained.

A.I.G

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

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (6)

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 94% confidence
Finding: The README explicitly advertises persistent cross-session memory and remembering information across sessions, but it does not warn users that potentially sensitive information may be retained beyond the current interaction. In an agent skill whose purpose is durable memory, this omission increases the risk that users or operators will store credentials, personal data, or other sensitive context without informed consent or retention controls.
Content: No source excerpt is available for this finding.

Undeclared Tool Scope

Medium

Category: MCP Least Privilege
Confidence: 84% confidence
Finding: The skill instructs the agent to read project memory files and invoke a local script, but it does not declare an explicit tool scope or permissions boundary. That mismatch can cause the runtime to grant broader-than-intended file access or make review of the skill's capabilities difficult, increasing the chance of unauthorized reads of local memory files.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The description says to use the skill whenever an .mneme file is present or whenever a user wants durable memory, which is broad enough to activate on many normal tasks. Over-triggering a persistent-memory skill increases the chance that routine conversations are treated as memory-worthy and that unrelated context is pulled into decision-making or storage.
Content: No source excerpt is available for this finding.

Ssd 3

Medium

Category: Not specified by scanner
Confidence: 94% confidence
Finding: This skill is specifically designed for persistent cross-session memory, but it provides no privacy scoping, sensitivity filtering, consent requirement, or retention minimization. In practice, that creates a clear path for storing secrets, personal data, credentials, or other sensitive user inputs durably in plaintext, where they may later be recalled or exposed.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 89% confidence
Finding: The 'When to use' section includes ambiguous triggers like 'You just produced something a future session would be wrong or redo without,' which leaves activation to agent discretion. That ambiguity can lead to over-collection, unnecessary file access, and memory persistence without a clear user signal or policy boundary.
Content: No source excerpt is available for this finding.

Ssd 3

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The write protocol tells the agent to append new durable memory whenever a future session might otherwise be wrong or redo work, which strongly encourages broad retention of user-provided information. Because the storage format is persistent plain text and the criteria are expansive, sensitive data can be captured unnecessarily and retained indefinitely across sessions.
Content: No source excerpt is available for this finding.