amazon-sorftime-research-keywords-skill

• Outcome: Pass
• Latest audit: May 29, 2026 · 8:28 PM UTC
• Version: 1.0.0

Overview

This skill is a disclosed Amazon keyword research workflow that sends product and keyword queries to Sorftime, uses optional LLM classification, and writes local reports; it has privacy and credential-handling caveats but no evidence of hidden or destructive behavior.

Install only if you are comfortable sharing ASINs, product details, and keyword research data with Sorftime and, when using LLM classification, with the selected LLM provider. Protect the Sorftime API key in .mcp.json, avoid logging full URLs that contain the key, and treat generated report files and classification prompts as potentially sensitive business research artifacts.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (8)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 95% confidence
• Finding: The skill describes and enables shell execution, environment access, and file read/write behavior, but does not declare permissions or clearly scope those capabilities. This creates a trust and review gap: users or platforms may not realize the skill can access local files, environment secrets, and execute commands that call external services.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The documentation instructs sending ASIN, product details, keyword data, and manual classification prompts to external Sorftime and Claude services, but provides no clear privacy notice, data retention statement, or handling guidance. This can lead to unintentional disclosure of business-sensitive research inputs and outputs to third parties without informed consent.

Missing User Warnings

Medium

• Confidence: 96% confidence
• Finding: The examples place the API key directly in the request URL query string. Query parameters are commonly exposed through shell history, process listings, logs, proxies, monitoring tools, and error reports, which increases the chance of credential leakage even when HTTPS is used. In this skill context, the document is operational guidance for real API use, so users are likely to copy the pattern verbatim.

Missing User Warnings

Medium

• Confidence: 98% confidence
• Finding: The documentation instructs callers to place the API key in the URL query string, which is commonly exposed in browser history, proxy logs, server logs, monitoring systems, and referrer-like telemetry. In this skill context, the pattern is more dangerous because the agent will likely automate outbound requests to a third-party MCP endpoint, increasing the chance that secrets are propagated into logs or diagnostics during normal operation.

Missing User Warnings

Medium

• Confidence: 88% confidence
• Finding: The workflow writes the full LLM classification prompt to disk, and that prompt includes product metadata plus up to 500 collected keywords. In a shared workspace, agent environment, or retained artifact store, this can expose potentially sensitive business research inputs and intermediate prompt content to other users, tools, or later processes without an explicit consent gate or minimization step.

External Transmission

Medium

• Category: Data Exfiltration

Content

#### Step 1.1: 获取产品流量词

```bash
curl -s -X POST "https://mcp.sorftime.com?key={API_KEY}" \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"product_traffic_terms","arguments":{"amzSite":"US","asin":"ASIN"}}}'
```

• Confidence: 92% confidence
• Finding: This step explicitly posts data to an external Sorftime endpoint using an API key in the request URL. Even though external transmission is core to the skill's function, it is still security-relevant because product and keyword research data leaves the local environment, and placing the API key in the URL increases the chance of leakage through logs, proxies, shell history, or monitoring systems.

External Transmission

Medium

• Category: Data Exfiltration

Content

### 调用格式

```bash
curl -s -X POST "https://mcp.sorftime.com?key={API_KEY}" \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":N,"method":"tools/call","params":{"name":"TOOL_NAME","arguments":{"amzSite":"US","KEY":"VALUE"}}}'
```

• Confidence: 91% confidence
• Finding: The generic call format normalizes a pattern of transmitting arbitrary tool arguments to an external endpoint and again embeds the API key in the URL. Because this is reusable documentation, it encourages broad outbound data flows and insecure secret handling across all supported tool calls, magnifying exposure if adopted as-is.

External Transmission

Medium

• Category: Data Exfiltration

Content

# Sorftime MCP 工具参考 (curl 调用格式)

**注意**：Sorftime MCP 使用 SSE 协议，所有工具调用格式如下：

• Confidence: 82% confidence
• Finding: The file instructs the agent/user to send product, review, keyword, and trend data to an external third-party service. External transmission is expected for an MCP integration, but it remains security-relevant because potentially sensitive business inputs and credentials are sent off-platform; combined with the query-string API key pattern, the risk is higher. In this skill context, the transmission is core functionality rather than covert exfiltration, so the issue is about data-handling exposure rather than malicious behavior.

VirusTotal

65/65 vendors flagged this skill as clean.

Static analysis

No suspicious patterns detected.