Back to skill

Security audit

API网关管理器(专业版)

Security checks for vulnerabilities and agentic risk

Overview

This skill is not clearly malicious, but its broad automation scope and inconsistent gateway-management versus grading/audit documentation make its high-impact exec/write authority risky to install without review.

Install only after confirming you want an agent to help manage API gateway configuration and after narrowing use to explicit gateway-management requests. Do not let it run apply, rollback, login, observability, or payment-log related commands against production without a reviewed config, a rollback plan, and explicit approval.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (6)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The activation guidance is extremely broad ('use when efficiency/automation/workflow optimization is needed'), which can cause the skill to trigger in many unrelated contexts. Because the skill has read/exec/write capabilities, overbroad routing increases the risk of unnecessary privileged actions or misuse outside the operator's intent.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The documentation includes commands that initialize dynamic control-plane behavior and apply/rollback gateway configuration, but it does not prominently require confirmation, staging, or impact acknowledgment before execution. In a skill with exec capability, users or downstream agents may treat these examples as safe defaults and cause disruptive production changes.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The plugin example forwards payment-related data, including highly sensitive fields such as card_number and cvv, to an external audit endpoint without a privacy or compliance warning. Even as an example, this normalizes transmission of PCI-regulated data and could lead users or agents to deploy configurations that exfiltrate sensitive payment information.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill presents itself as an API gateway management tool, but its documented outputs and schemas describe a generic grading/audit engine. This mismatch can mislead users and downstream agents into invoking the skill on inappropriate inputs or trusting outputs for actions unrelated to gateway management, increasing the chance of unsafe automation and incorrect operational changes.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The inline documentation continues to claim gateway-management functionality while examples show a generic scoring workflow. In an agentic environment, contradictory docs can cause the model to improvise behavior, potentially issuing file, exec, or config-changing actions under false assumptions about what the skill is supposed to do.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

The description says the skill 'supports Chinese interaction' as a fixed behavior and does not indicate that users may choose another language. Under the locale policy, forcing or defaulting to a specific language without opt-in is a natural-language policy issue unless the regional constraint is justified and documented.

Content

No source excerpt is available for this finding.