Security audit

Amap LBS Service

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly aligned with map/search use, but it automatically sends usage pings, embeds user data URLs into an Amap page, and ships with an apparent API key in config.

Review this skill before installing. Use it only for map-related tasks, avoid providing private or signed data URLs for heatmaps, and replace or remove the bundled Amap key so requests use an account you control. Expect external requests to Amap, including usage-count telemetry before operations.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
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

Findings (6)

Context-Inappropriate Capability

Medium

Confidence: 96% confidence
Finding: The skill explicitly instructs sending a telemetry request before performing a simple keyword search, but that network call is not required to fulfill the user’s requested map search. This causes unnecessary disclosure of request metadata to a third party and creates hidden external side effects beyond the stated user task.

Context-Inappropriate Capability

Medium

Confidence: 97% confidence
Finding: The nearby-search flow adds an analytics request before geocoding and link generation, even though it is unrelated to producing the nearby search result. Because nearby searches may encode sensitive location intent, this extra call increases privacy risk by transmitting operation metadata to Amap without necessity.

Context-Inappropriate Capability

Medium

Confidence: 95% confidence
Finding: The heatmap workflow includes a telemetry request that is not needed to generate the visualization URL. In a flow that already handles user-supplied data locations, adding another unsolicited outbound request compounds privacy exposure and creates unnecessary hidden behavior.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The trigger conditions include very broad common terms such as “搜”, “找”, “查”, “附近”, and “规划”, which can match many unrelated user requests. Overbroad activation can cause the wrong skill to run, leading to unintended external requests, unnecessary data sharing, or confusion in security-sensitive contexts.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: Although the note section later mentions anonymous usage statistics, the operational instructions themselves direct telemetry on each execution without clear, upfront user-facing consent at the point of action. Users may reasonably believe the skill only generates a search link, while in reality it also transmits request metadata externally.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The heatmap feature accepts arbitrary user-provided data URLs and embeds them into an Amap-hosted visualization page without a strong privacy warning. This can expose sensitive datasets, signed URLs, internal endpoints, or access tokens to external infrastructure when users may not realize their data source will be fetched or revealed by a third-party page.

Static analysis

No suspicious patterns detected.