Back to skill

Security audit

导出治理与审批 (Agentic AI 科研平台)

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent export-governance helper that reads a fixed governance data file and sends disclosed local status updates, with some scoping and local-service assumptions users should understand.

Install only in an environment where the hardcoded governance JSON path and localhost reporting service are expected. Treat the displayed governance data as potentially sensitive, and avoid using the broad approval triggers outside export-governance tasks.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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

示例触发语包括“帮我导出分析结果”“提交审批”“看一下脱敏规则”,其中“提交审批”和“看一下脱敏规则”缺少上下文约束,可能出现在多种普通对话或其他工作流中。文档虽说明用于“单独执行”,但没有给出明确的边界条件或排除示例,触发范围仍然偏模糊。

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
89% confidence
Finding

The skill instructs posting status messages to an HTTP endpoint via curl, which is an external transmission channel and uses plaintext transport. Even though the payload shown is operational metadata rather than research data, this creates a path for unauthorized disclosure, tampering, or abuse if the local service is exposed, compromised, or repurposed; in a governance/export context, any unreviewed outbound reporting is more sensitive because it accompanies data-export workflows.

Content

Scanner excerpt · SKILL.md (reported line 22)May include surrounding context.

1. 上报开始

bash
curl -s -X POST http://localhost:5001/api/report \
  -H "Content-Type: application/json" \
  -d '{"skill":"export-governance","display_name":"导出治理与审批","status":"running","message":"正在识别导出风险,应用脱敏规则..."}'