Security audit

MapV Three（百度地图官方3D地图SKills）

Security checks for vulnerabilities and agentic risk

Overview

This is a documentation-only MapV-Three skill whose API-key and map-service guidance fits its GIS purpose, but users should sanitize copied popup/DOM examples and restrict map credentials.

Install with normal caution for a map-development reference skill: use restricted and rotatable Baidu/Mapbox/Cesium/Tianditu keys, avoid committing production credentials, review provider privacy and billing terms, and do not copy raw innerHTML or popup HTML examples for untrusted user or dataset content without escaping or sanitizing it first.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (5)

Context-Inappropriate Capability

Medium

Confidence: 97% confidence
Finding: The DOMPoint example assigns untrusted data fields directly into `node.innerHTML`, which can enable DOM-based XSS if `name` or `description` contain attacker-controlled HTML or script gadgets. In a Web-GIS application, point metadata often comes from external datasets, APIs, or user input, so this documentation pattern can easily propagate into real applications and lead to script execution in the browser.

Missing User Warnings

Low

Confidence: 94% confidence
Finding: The skill requires an API credential via BMAP_JSAPI_KEY but does not disclose how the key is used, whether requests go to third-party services, or what privacy/account-billing implications exist. In an agent ecosystem, undisclosed credential requirements can mislead users into exposing secrets to tooling they do not fully understand, even if the key is expected for legitimate map service access.

Missing User Warnings

Medium

Confidence: 98% confidence
Finding: The sample code demonstrates unsafe HTML rendering without any warning, normalizing an insecure pattern for users of the library. Because this is reference documentation for a reusable GIS component, developers are likely to copy the example verbatim, turning untrusted map feature properties into executable markup and creating stored or reflected XSS paths in consuming applications.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The documentation explicitly shows assigning raw HTML strings to `popup.content` and `popup.title` without any warning or sanitization guidance. If developers pass untrusted or user-controlled data into these properties, this can lead to DOM-based XSS in the consuming application, especially because this is a popup/UI component intended to render arbitrary markup.

Missing User Warnings

Medium

Confidence: 89% confidence
Finding: The document describes geocoding, POI search, autocomplete, and route-planning features that inherently send user-provided addresses, keywords, and location coordinates to external map providers such as Baidu Maps or Tianditu, but it does not clearly disclose that third-party data transfer occurs. In a GIS/LBS context, these queries can contain sensitive location or intent information, so lack of transparency can lead to privacy, consent, and compliance issues even if the behavior is expected for the feature.

Static analysis

No suspicious patterns detected.