Back to skill

Security audit

MCM Agent

Security checks for vulnerabilities and agentic risk

Overview

This is a small cloud-memory helper skill that discloses remote memory syncing and consent requirements, though its activation triggers are broader than ideal.

Before installing, be comfortable giving this service an API key and storing selected memory content in its cloud account. Use explicit save or recall requests, and avoid relying on it for sensitive secrets unless you trust the provider and its retention controls.

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
  • 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 (3)

Vague Triggers

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The trigger list is broad enough to activate on ordinary discussion of memory, syncing, or related Chinese terms, including very short generic words. Because the skill is marked user-invocable: false, these triggers may cause passive invocation without a clear user request, increasing the chance that memory-related behavior is introduced unexpectedly. In a skill that can save or recall cloud-stored user information, accidental activation raises privacy and consent risks.

Content

No source excerpt is available for this finding.

Overly Broad Trigger

Low
Category
Trigger Abuse
Confidence
99% confidence
Finding

The trigger '记忆' is a very common standalone word that can appear in many benign conversations, making unintended skill activation likely. In this context, the skill handles cloud memory synchronization and retrieval, so even if later consent is requested, unexpected invocation can expose prior stored memory context or pressure users into unintended data operations.

Content

No source excerpt is available for this finding.

Overly Broad Trigger

Low
Category
Trigger Abuse
Confidence
99% confidence
Finding

The trigger '同步' is overly generic and likely to match many unrelated conversations about synchronization. Given that this skill interfaces with a remote cloud memory account, accidental activation can create privacy and workflow risks by surfacing or attempting to act on stored user data in contexts where the user did not intend to invoke the skill.

Content

No source excerpt is available for this finding.