Back to skill

Security audit

Agent Self Evaluation

Security checks for vulnerabilities and agentic risk

Overview

This skill provides a self-evaluation checklist for agent outputs and does not include code execution, credential access, persistence, or data exfiltration behavior.

This skill is reasonable to install if you want structured self-checks after substantial work. Expect it may add extra review text if triggered broadly; publishers should narrow the description so it runs mainly for complex tasks or explicit self-evaluation requests.

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 (2)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill description says it should be used when completing any task, which makes invocation scope extremely broad. Overly broad triggers can cause unnecessary self-evaluation to run in many unrelated contexts, increasing prompt bloat, context pollution, and the chance that sensitive or adversarial content gets reprocessed when it is not needed.

Content

No source excerpt is available for this finding.

Vague Triggers

Low
Category
Not specified by scanner
Confidence
86% confidence
Finding

The example user phrase for invoking the skill is generic enough to overlap with ordinary conversation, which can lead to accidental activation. That can cause the agent to switch into an evaluation workflow when the user intended a casual question, reducing reliability and potentially exposing prior task context in an unnecessary self-assessment.

Content

No source excerpt is available for this finding.