Taxi

• Outcome: Pass
• Latest audit: May 28, 2026, 12:55 PM
• Version:
  1.0.0

Overview

This taxi helper is purpose-aligned and instruction-only, but it can save sensitive ride and location details locally and should be supervised during bookings.

Install only if you are comfortable with a local ~/taxi folder that may contain home/work addresses, trip history, receipts, account identifiers, and promo details. Avoid saving passwords, full payment details, or exact addresses unless necessary; review booking details yourself before confirming any ride or payment; and keep account-specific complaints or receipts out of public social posts.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• 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 (7)

Vague Triggers

Medium

• Confidence: 92% confidence
• Finding: The invocation scope is extremely broad ('ANY help with taxis, rideshares, car services, or ground transportation'), which can cause the skill to activate for loosely related travel queries and perform actions in contexts the user did not specifically intend. In a skill that can open booking sites, pre-fill rides, and use persisted personal travel context, overbroad triggering increases the chance of privacy-invasive behavior or unintended transactional assistance.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The skill persistently stores sensitive travel data in local files, including saved addresses, trip history, expenses, app accounts, and promos, but does not present an explicit warning, consent flow, retention policy, or data minimization guidance. This creates meaningful privacy and security risk because home/work locations, routines, business travel, and account metadata can expose a user's habits and physical safety if accessed by other local users, malware, or over-permissive tooling.

Missing User Warnings

Medium

• Confidence: 88% confidence
• Finding: The skill instructs the agent to navigate third-party ride-booking flows that inherently involve sensitive data such as precise pickup/destination locations, account sessions, and payment-adjacent screens, but it does not include explicit privacy, consent, or data-handling safeguards. In this context, the omission increases the risk of exposing sensitive travel patterns or mishandling authenticated sessions even though the final payment click is reserved for the user.

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The skill explicitly instructs storing a user's common routes and historical trip prices in `~/taxi/trips.md`, which can reveal sensitive travel patterns, home/work locations, routines, and event attendance. In a taxi/transportation skill, this data is especially privacy-sensitive, and the file path suggests persistent local storage without any notice, consent, retention limit, or access-control guidance.

Missing User Warnings

Medium

• Confidence: 93% confidence
• Finding: The template explicitly encourages storing sensitive personal and account data locally, including home/work addresses, phone numbers, emails, payment preferences, trip history, and referral/account details, but provides no warning about privacy, device security, or file protection. In a transportation skill, this creates a meaningful privacy and profiling risk if the local files are exposed through device compromise, backups, shared accounts, or other skills reading the same directory.

Missing User Warnings

Medium

• Confidence: 89% confidence
• Finding: The skill advises users to escalate through public social media for support, but it does not warn them to avoid posting trip IDs, pickup/dropoff locations, receipts, or other personal details publicly. In a transportation context, those details can expose sensitive location history, account information, and safety-related incident data to a broad audience, increasing privacy and stalking risks.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The documentation template instructs users to compile detailed trip metadata including date, trip ID, pickup and destination addresses, fare details, and attachments, but provides no privacy handling guidance. In a taxi support skill, this information is highly sensitive because it reveals travel patterns, home/work locations, and potentially financial data that could be mishandled, overshared, or copied into insecure channels.

VirusTotal

66/66 vendors flagged this skill as clean.