fastapi-streamlit

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill provides templates for creating applications that ingest untrusted data from external sources.
  • Ingestion points: Web API endpoints (@app.post) and user file uploaders (st.file_uploader) documented in SKILL.md.
  • Boundary markers: Absent, as the templates focus on standard web development protocols rather than LLM-specific prompt engineering.
  • Capability inventory: The templates enable data processing, file reading, and model inference via Python scripts.
  • Sanitization: The skill strongly encourages the use of Pydantic schemas for strict type validation and input sanitization.
  • [DYNAMIC_EXECUTION]: The skill demonstrates standard methods for loading machine learning models.
  • Evidence: The 'Scientific Model Serving' example in SKILL.md utilizes torch.load("model.pth").
  • Note: While torch.load involves deserialization, it is a standard practice for the skill's primary purpose of model deployment; users should ensure models come from trusted sources.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 04:41 PM
Security Audit — agent-trust-hub — fastapi-streamlit