rust-web
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to handle untrusted data through HTTP request extractors (Path, Query, Header, Body). It includes mandatory security guidance to mitigate injection and exposure risks:
- Ingestion points: Data enters the system via Axum extractors as defined in the routing logic.
- Boundary markers: The skill enforces a strict layered architecture (Transport -> Service -> Domain -> Repository) to isolate transport-layer data from business logic.
- Capability inventory: The skill outlines capabilities for database access (SQLx) and outbound HTTP requests (reqwest).
- Sanitization: Explicit instructions are provided to redact sensitive headers, mask internal error details (SQL queries, stack traces) from responses, and exclude secrets from logging streams.
- [COMMAND_EXECUTION]: The skill includes standard local development commands (
cargo test,cargo clippy,cargo fmt) used for project maintenance and verification. These do not involve downloading or executing untrusted remote scripts.
Audit Metadata