Back to skill

Security audit

HR ArchSystem

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent HR framework, but it should be reviewed because it guides sensitive employee evaluations using personnel, feedback, performance, and AI-use data without privacy or retention safeguards.

Install only for authorized HR or management use. Before using it on real employees, define who may provide and see the data, whether employee notice or consent is required, what AI/tool logs may be used, how reports are redacted, and when evaluation records are deleted or reviewed.

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 (6)

Missing User Warnings

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

The document explicitly directs collection of sensitive personnel information, behavioral evidence, peer/manager feedback, and AI usage logs, but provides no privacy notice, lawful-basis guidance, retention limits, access controls, or confidentiality handling requirements. In an HR evaluation context this is especially dangerous because the data is highly sensitive, can affect promotion and compensation decisions, and could create privacy, insider-risk, and compliance exposure if gathered or shared improperly.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The trigger conditions are broad natural-language phrases such as requests to define capability requirements or measure a person's ability level, without clear scoping or exclusions. This can cause the skill to activate on adjacent HR, performance, or general advice queries unexpectedly, leading to over-collection of personnel context or routing sensitive evaluation tasks into this module when a narrower workflow should apply.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The trigger phrases are broad enough to match ordinary HR questions such as evaluating someone’s level or writing qualification materials, which can cause the skill to activate when the user did not explicitly intend to invoke this evaluation workflow. In this skill’s context, unintended activation is more dangerous because the workflow is designed to collect and structure sensitive personnel, performance, and AI-usage evidence, so accidental use could expose or process confidential employee data without clear consent or scoping.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

This markdown skill content forces a specific language for all users, which can violate language or locale policy when no opt-in or alternative is provided. The file contains no indication that the skill is China-specific or that users may choose another language.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger condition includes a very broad natural-language phrase ('我们应该怎么组织' 类似问题), which can cause the skill to activate during ordinary discussion rather than only for explicit HR architecture tasks. In an agent system, over-broad activation increases the chance of unintended context capture, inappropriate tool/skill invocation, and downstream handling of sensitive organizational or personnel information without clear user intent.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

This skill content appears to require Chinese by default, and there is no natural-language statement offering users a language option or explaining that the skill is intentionally limited to a Chinese-speaking context. Under the language/locale policy, forcing a specific language without opt-in can be a policy issue.

Content

No source excerpt is available for this finding.