litestar-app-setup
Fail
Audited by Gen Agent Trust Hub on Mar 7, 2026
Risk Level: HIGHCREDENTIALS_UNSAFEPROMPT_INJECTIONSAFE
Full Analysis
- [CREDENTIALS_UNSAFE]: Hardcoded database credentials detected in code examples within
SKILL.md.\n - Evidence: The connection string
postgresql+asyncpg://postgres:mysecretpassword@pg.db:5432/dbcontaining a plaintext password is used in the 'Startup and Shutdown Hooks' and 'Lifespan Context Managers' sections.\n- [PROMPT_INJECTION]: Vulnerability to indirect prompt injection through processing untrusted user task descriptions.\n - Ingestion points: User instructions for Litestar tasks provided to the agent as defined in
agents/openai.yaml.\n - Boundary markers: No delimiters or instructions to ignore embedded commands are present in the prompt templates.\n
- Capability inventory: The skill creates Python application code, configures database engines, and implements lifecycle hooks.\n
- Sanitization: No validation or escaping mechanisms are described for user-provided input before it is used to generate application configuration.\n- [SAFE]: External documentation references point to trusted official sources.\n
- Evidence: Links in the 'Litestar References' section of
SKILL.mdcorrectly point to the officialdocs.litestar.devdocumentation.
Recommendations
- AI detected serious security threats
Audit Metadata