Back to skill

Security audit

EO Ability: Memory

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed memory helper skill with no bundled executable code or install steps, though users should treat its automatic cross-session memory as privacy-sensitive.

Before installing, confirm whether EO memory is enabled, where persistent memories and LanceDB indexes are stored, and how to review, delete, disable, or scope them by project. Avoid saving secrets, regulated data, credentials, or confidential negotiations unless you have explicit controls for retention and access.

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
  • 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 (3)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill explicitly advertises automatic cross-session saving and restoration of user preferences, mentor requirements, and project context, but provides no consent, minimization, or privacy boundary guidance. Because this capability is designed to retain potentially sensitive personal and project information across sessions, users may unknowingly persist confidential data and have it resurfaced later in unrelated contexts.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The examples normalize EO automatically triggering save/load actions without any warning that sensitive information may be persisted or automatically reintroduced into future conversations. This increases the likelihood of inadvertent retention and disclosure of confidential preferences, research details, or business negotiation context, especially because the behavior is framed as default automation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
77% confidence
Finding

Nearly all user-facing instructional content is in Chinese, with no indication that other languages are supported or that Chinese is required for a specific regional use case. This may constitute a language/locale policy issue if the organization expects skills not to force a language without user opt-in.

Content

No source excerpt is available for this finding.