Safe Driving Behavior Analyzer | 安全驾驶行为分析工具

• Outcome: Review
• Latest audit: Aug 27, 2026 · 6:51 AM UTC
• Version: 1.0.11

Overview

The skill mostly matches its driving-video analysis purpose, but it also performs under-disclosed identity, account, token, local database, and remote API behavior that users should review before installing.

Review this skill carefully before installing. It can upload driving videos or URLs for remote analysis, query cloud-stored report history, read or create a local identity, store tokens in a workspace SQLite database, and contact configured service endpoints. Install only if you trust the publisher and are comfortable with identity-linked media analysis and local token persistence.

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
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (9)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 95% confidence
• Finding: The skill instructs the agent to run local Python scripts, save uploaded files locally, access cloud APIs, and potentially use environment/configuration data, yet it declares no permissions. This creates a capability-transparency gap: users and orchestrators are not clearly warned that the skill can read/write files, invoke shell commands, and send data over the network.

Vague Triggers

Medium

• Confidence: 88% confidence
• Finding: The automatic trigger phrases for historical report retrieval are broad enough to match ordinary conversational requests, which can cause unintended execution of the cloud-history query path. In this skill, that means potentially sending or retrieving user-linked report data without sufficiently explicit intent confirmation.

Missing User Warnings

Medium

• Confidence: 93% confidence
• Finding: The skill states that uploaded attachments and video/image files will be automatically saved as local files, but it does not present this as a clear user warning or consent point. This is dangerous because uploaded media may contain sensitive personal data, and silent local persistence increases privacy and retention risks.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The skill requires history-report queries to be performed through a cloud API, but the description does not clearly warn users that their requests and associated report data are sent to a remote service. Because the reports are identity-linked internally, this omission creates a meaningful privacy and transparency issue.

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: When a local path is provided, the skill reads the entire file and sends its contents to a remote analysis API, but this code shows no visible consent prompt, warning, or data-handling notice at the point of upload. Because driving videos can contain sensitive personal data such as faces, license plates, location cues, and bystanders, silent transmission increases privacy and compliance risk if users do not realize the file leaves the local environment.

Missing User Warnings

Medium

• Confidence: 96% confidence
• Finding: The code silently reads an identifier from data/smyx-api-key.txt and uses it as an internal identity value without user notice or consent. In an agent skill context, quietly harvesting credentials or identity material from workspace files can surprise users, cross trust boundaries, and enable unintended account association or impersonation.

Missing User Warnings

Medium

• Confidence: 98% confidence
• Finding: This request wrapper automatically attaches tokens, API keys, tenant metadata, and usernames to outbound HTTP requests, while also performing automatic authentication flows. In a skill environment, hidden transmission of identity and auth material to remote services is dangerous because it can leak credentials, create unanticipated remote side effects, and make user actions non-transparent.

Missing User Warnings

Medium

• Confidence: 98% confidence
• Finding: The helper silently performs phoneLogin with silent=1 and register=1, meaning it may create or log into an account over the network without explicit user approval. That is especially risky in an agent skill because merely invoking functionality can trigger account provisioning and identity linkage to a remote service without the user understanding it.

External Transmission

Medium

• Category: Data Exfiltration

Content

"source": ConstantEnum.DEFAULT__SKILL_HUB_NAME
            }
            try:
                _response = requests.post(_url, json=_data)
                if _response.status_code == 200:
                    _response_json = _response.json()
                    if _response_json and _response_json.get("success"):

• Confidence: 97% confidence
• Finding: requests.post(_url, json=

VirusTotal

VirusTotal findings are pending for this skill version.

Static analysis

Detected: suspicious.install_untrusted_source

Install source points to URL shortener or raw IP.

Warn

• Code: suspicious.install_untrusted_source
• Location: skills/smyx_common/scripts/config-dev.yaml:2