helix-query-rust

Pass

Audited by Gen Agent Trust Hub on Jun 26, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides documentation, canonical examples, and an exhaustive DSL reference for the helix-db Rust SDK. It contains no executable scripts, persistence mechanisms, or obfuscated code. All network-related examples correctly utilize placeholders for API keys and endpoint URLs.
  • [PROMPT_INJECTION]: The skill describes an interface for processing external data via database query parameters, which represents an indirect prompt injection surface.
  • Ingestion points: Untrusted data enters the query context through parameters such as userId, tenantId, and queryVector as documented in SKILL.md and EXAMPLES.md.
  • Boundary markers: The DSL enforces strong boundaries using methods like Predicate::eq_param and PropertyInput::param, which ensure data is treated as a literal value rather than executable query logic.
  • Capability inventory: The skill facilitates database read/write operations and network communication with the database server using the Client struct.
  • Sanitization: The SDK's DSL architecture enforces parameterization at the AST level, providing built-in protection against query injection attacks.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 26, 2026, 08:10 PM
Security Audit — agent-trust-hub — helix-query-rust