fastapi-api-patterns

Warn

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: MEDIUMREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • Dynamic Execution (MEDIUM): The script scripts/generate-openapi-docs.sh uses importlib.import_module and getattr to load a FastAPI application from a module path provided as a command-line argument. This allows for arbitrary code execution during the module import phase if the script is directed to process an untrusted or malicious module.
  • Indirect Prompt Injection (MEDIUM): The skill exhibits an attack surface (Category 8) because scripts/validate-endpoints.sh and scripts/generate-openapi-docs.sh ingest untrusted data from files provided as arguments (Ingestion points: $1 filename argument in both scripts). There are no boundary markers or sanitization routines present to prevent malicious content in those files from influencing script output or agent reasoning (Boundary/Sanitization: Absent). The scripts possess significant system capabilities including Python execution and shell command processing (Capability inventory: subprocess calls, import_module, grep, sed).
  • External Downloads (LOW): The scripts/generate-openapi-docs.sh script generates HTML documentation that includes remote script and stylesheet links to cdn.jsdelivr.net. While these are standard Swagger UI libraries, they represent a reliance on external third-party assets.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 16, 2026, 05:35 AM