Security audit

DB Travel

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward Deutsche Bahn travel helper that makes disclosed public transit API lookups and does not show hidden persistence, credential access, or destructive behavior.

Before installing, be aware that travel searches, station names, stop IDs, and requested journey times are sent to the public DB transport REST API. The skill does not buy tickets or handle credentials, and its English-only API parameter may limit localized results.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration

Findings (6)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 91% confidence
Finding: The skill invokes a network-capable script against an external public API, but the manifest does not declare any corresponding tool scope or permission boundary. This weakens reviewability and policy enforcement because users and orchestrators cannot easily tell that activating the skill may cause outbound network requests and data disclosure of user-provided travel queries.

Vague Triggers

Medium

Confidence: 94% confidence
Finding: The manifest says the skill 'triggers on questions about German trains, DB, BVG, getting around Berlin, European rail connections,' which is broad and overlaps with many ordinary travel questions. It does not define clear boundaries or exclusion conditions for when this skill should activate versus other transit or travel skills.

Natural-Language Policy Violations

Medium

Confidence: 94% confidence
Finding: The code hard-codes the API parameter `language` to `en`, which imposes a specific language on returned journey data. The file does not offer user opt-in or language selection, and no region-specific justification is documented for this constraint.

Natural-Language Policy Violations

Medium

Confidence: 94% confidence
Finding: The departures command always sends `language: en` to the backend, enforcing an English locale choice in natural-language output. This is a policy concern because users are not given any way to choose or consent to the language setting.

Natural-Language Policy Violations

Medium

Confidence: 94% confidence
Finding: The arrivals command hard-codes `language` to `en`, which forces a specific locale for response text. No user choice or documented rationale is provided in the file.

Natural-Language Policy Violations

Medium

Confidence: 93% confidence
Finding: The stop lookup endpoint is called with `language: en`, imposing English output regardless of user preference. The policy allows locale constraints only when users can opt in or the restriction is clearly justified.