SerpAPI

• Outcome: Pass
• Latest audit: May 27, 2026, 9:12 PM
• Version:
  1.0.0

Overview

This is a straightforward SerpAPI search wrapper that uses a configured API key and sends searches to SerpAPI, with privacy notes but no evidence of hidden or malicious behavior.

Install only if you are comfortable sending search terms, optional location data, and SerpAPI account usage to SerpAPI. Use a dedicated or revocable SERPAPI_API_KEY if possible, monitor quota usage, and avoid configuring a default location in TOOLS.md if you do not want location context reused automatically.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration

Findings (4)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 88% confidence
• Finding: The skill documentation indicates access to environment variables, file reads, and network use via metadata and setup instructions, but it does not declare corresponding permissions. That mismatch weakens reviewability and consent because an agent may use capabilities that are not transparently surfaced to operators, especially with access to `SERPAPI_API_KEY` and local configuration files like `TOOLS.md`.

Context-Inappropriate Capability

Medium

• Confidence: 93% confidence
• Finding: The skill reads local workspace files (including a path derived from an environment variable) to infer a default location, even though its primary purpose is external search. This creates unnecessary access to local context and can leak sensitive or user-specific location data to the external SerpAPI service without an explicit user request.

Vague Triggers

Medium

• Confidence: 76% confidence
• Finding: The description uses broad trigger language such as general web search, products, local businesses, restaurants, shopping, images, and news, making the skill likely to be invoked for many ordinary user requests. In a networked skill, over-broad invocation increases the chance of unnecessary external requests and unintended disclosure of user queries or location data to a third-party service.

Missing User Warnings

Medium

• Confidence: 88% confidence
• Finding: The code sends user queries and optional location data to a third-party API endpoint, but there is no explicit warning or consent flow informing the user that their inputs will leave the local environment. In a search skill, network transmission is expected, but silently forwarding inferred location data increases privacy risk and can surprise users.

VirusTotal

66/66 vendors flagged this skill as clean.