fastapi-patterns
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides patterns for building web APIs that ingest untrusted user data, creating a potential surface for indirect prompt injection.
- Ingestion points: Data enters through the
SkillCreatePydantic model in thecreate_skillendpoint and via query parameters in thelist_skillsendpoint inSKILL.md. - Boundary markers: The skill demonstrates the use of Pydantic models for schema validation, which provides structural boundaries but does not automatically sanitize natural language instructions within strings.
- Capability inventory: The defined patterns include database write operations (
db.add,db.commit) and the execution of background tasks viaBackgroundTasksinSKILL.md. - Sanitization: Validation is limited to structural and type checks provided by Pydantic; no explicit sanitization for LLM-directed instructions is demonstrated.
Audit Metadata