Back to skill

Security audit

medical_invoice_scene_ocr

Security checks across malware telemetry and agentic risk

Overview

This skill appears to be a legitimate Scnet medical OCR wrapper, but it can upload sensitive medical and identity documents to a third-party API without strong consent or privacy guidance.

Install only if you intend to send medical reimbursement documents to Scnet's external OCR service and have authority to do so. Do not use it for regulated patient, employee, or customer records unless you have reviewed Scnet's privacy, retention, deletion, residency, and compliance terms and can ensure users explicitly approve each upload.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (7)

Lp3

Medium
Category
MCP Least Privilege
Confidence
87% confidence
Finding
The skill documentation indicates capabilities to read local files, invoke Python from the shell, and send data over the network, but it does not declare permissions or clearly bound those actions. This creates a real security gap because an orchestrator or reviewer cannot enforce least privilege, while the skill processes sensitive medical images and transmits them to a remote API.

Vague Triggers

Medium
Confidence
81% confidence
Finding
The activation guidance is broad enough that an AI agent may auto-trigger this skill whenever a user mentions a medical invoice image and a path, without strong confirmation that remote OCR and local file access are intended. In this context, that can cause unintended transmission of highly sensitive medical billing documents to a third-party service.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The documentation instructs clients to upload medical invoices and settlement documents containing highly sensitive personal and health data to a third-party OCR endpoint, but it does not warn about the privacy, compliance, retention, or cross-boundary transmission risks. In this medical reimbursement context, the example payloads include names, ID numbers, insurance numbers, hospitalization details, and diagnoses, making omission of data-handling guidance materially dangerous.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The script uploads medical invoice images to a third-party OCR API, which likely contains sensitive personal and health-related data. While the transmission appears intentional for functionality, there is no explicit user-facing consent or warning at the moment of upload, creating a privacy and compliance risk that is elevated in a medical reimbursement context.

External Transmission

Medium
Category
Data Exfiltration
Content
| 变量名 | 默认值 | 说明 |
|--------|--------|------|
| SCNET_API_KEY | 必需 | Scnet API 密钥 |
| SCNET_API_BASE | https://api.scnet.cn/api/llm/v1 | API 基础地址(一般无需修改) |

### 输出
Confidence
91% confidence
Finding
https://api.scnet.cn/

External Transmission

Medium
Category
Data Exfiltration
Content
# Sugon-Scnet OCR API 文档摘要

## 接口地址
`POST https://api.scnet.cn/api/llm/v1/ocr/recognize`

## 请求头
- `Content-Type: multipart/form-data`
Confidence
93% confidence
Finding
https://api.scnet.cn/

External Transmission

Medium
Category
Data Exfiltration
Content
)
        sys.exit(error_msg)

    config.setdefault('SCNET_API_BASE', 'https://api.scnet.cn/api/llm/v1')
    return config

def recognize_with_retry(ocr_type, file_path, config, retry_count=0):
Confidence
91% confidence
Finding
https://api.scnet.cn/

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.