openai-ads-api
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [SAFE]: The skill implements several security best practices, including explicit instructions to redact bearer tokens from output, requirements for restrictive file permissions (chmod 600) on local credentials, and mandatory user confirmation before performing actions that could increase advertising spend.
- [COMMAND_EXECUTION]: The skill provides bash code snippets for the agent to securely manage API tokens, such as using
read -rsfor silent input andchmodfor setting directory and file permissions. - [INDIRECT_PROMPT_INJECTION]: The skill processes external data from the OpenAI Ads API which represents a standard indirect prompt injection surface. 1. Ingestion points: API responses from
api.ads.openai.comcontaining user-controllable fields like campaign names and ad creative text. 2. Boundary markers: The skill does not define explicit delimiters or 'ignore' instructions for the content of ad fields. 3. Capability inventory: Network access viacurlfor API mutations and file system access for token management. 4. Sanitization: The skill mandates the redaction of bearer tokens and other secrets from API failure reports.
Audit Metadata