analytics-engineer

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:43 PM UTC
• Version:
  1.0.0

Overview

This is a documentation-only analytics engineering skill, but users should treat its runnable examples as production-sensitive templates.

Install this only if you want an analytics engineering reference skill. Before using generated code from it, move credentials into a secrets manager, use least-privilege service accounts, test in development or staging, and require explicit approval before refreshing BI assets, deploying dbt to production, querying sensitive warehouse data, or sending email alerts.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• 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 (5)

Context-Inappropriate Capability

Medium

• Confidence: 89% confidence
• Finding: The example includes authenticated calls to an external BI platform and the ability to trigger remote datasource refreshes, which introduces operational side effects beyond passive analytics examples. In a reference/examples file, this can normalize network-capable automation and encourage users to run code that reaches external services with privileged credentials.

Context-Inappropriate Capability

Medium

• Confidence: 84% confidence
• Finding: This code couples local command execution with downstream BI refresh automation, creating a workflow that can execute system commands and then trigger external state changes. Even though the commands are fixed, the pattern expands the skill from analytics modeling into orchestration with side effects, increasing the risk of unsafe reuse in agent contexts.

Context-Inappropriate Capability

Medium

• Confidence: 87% confidence
• Finding: The example demonstrates direct warehouse connectivity plus outbound email alerting, which goes beyond transformation/modeling guidance and introduces data access and exfiltration-capable behaviors. In an agent skill, such examples can be repurposed to query sensitive data and send alerts or data externally without clear guardrails.

Missing User Warnings

Medium

• Confidence: 98% confidence
• Finding: The example hardcodes a service username and password directly in source, which is an insecure secret-handling practice. Embedded credentials are frequently copied into real deployments, exposed in repositories, logs, or screenshots, and can enable unauthorized access to BI infrastructure.

Missing User Warnings

Medium

• Confidence: 98% confidence
• Finding: The example includes database credentials in code and demonstrates outbound email behavior without warning, combining secret exposure with a potential external transmission path. This pattern is especially risky in documentation because users often copy examples verbatim into production scripts.

VirusTotal

63/63 vendors flagged this skill as clean.