openai-ads-conversions-setup
Pass
Audited by Gen Agent Trust Hub on Aug 28, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [Trusted Vendor Content]: The skill is authored by a known vendor and utilizes official domains (openai.com) and delivery networks (bzrcdn.openai.com) for its SDK and documentation. These resources are integral to the skill's intended functionality.
- [Credential Protection]: The instructions emphasize the prevention of secret exposure. It provides specific scripts,
verify_capi_secret_not_exposed.pyandverify_ads_setup.py, designed to perform static analysis of the repository to detect and prevent the hardcoding of API keys or secrets. - [External Resource Integration]: The skill integrates the OpenAI Ads Measurement Pixel by loading a JavaScript SDK from
https://bzrcdn.openai.com/sdk/oaiq.min.js. This is a standard implementation pattern for browser-side analytics and originates from the vendor's own infrastructure. - [Local Command Execution]: The workflow involves running localized Python scripts for verification. These scripts use standard libraries to perform pattern matching on repository files and do not exhibit suspicious behavior such as network exfiltration or dynamic code execution of untrusted input.
- [Data Ingestion Surface]: The skill analyzes repository code to identify frameworks and existing instrumentation. While this involves reading untrusted data (the repository contents), the process is governed by strict instructional rules to prioritize official documentation and existing patterns, reducing the potential impact of indirect prompt injection.
Audit Metadata