logging-api-requests
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill facilitates the implementation of security best practices, specifically PII redaction and structured audit logging. All external references target well-known services or the author's own infrastructure.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for processing untrusted data from API request bodies and headers. While this creates a potential surface for injection, the skill includes explicit instructions for sanitization and redaction to mitigate these risks.
- Ingestion points: HTTP request data (bodies, headers, paths) processed by the generated middleware, and existing project files accessed via Read and Grep tools.
- Boundary markers: None explicitly defined in the agent's instructions for analyzing existing code.
- Capability inventory: The skill utilizes Write for file creation and a scoped Bash tool for boilerplate generation.
- Sanitization: Explicit instructions for PII redaction using configurable patterns and mandatory structured JSON serialization are included.
Audit Metadata