test-quality
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and follow instructions from external configuration and documentation files within the user's repository.
- Ingestion Points: The skill reads
.lattice/config.yamlfrom the repository root and subsequently reads files pointed to bypaths.test_qualityandpaths.language_idioms. - Capability Inventory: The agent is instructed to use the content of these external files to override its own logic (
mode: override), add new validation checks, and adapt its code generation patterns. - Boundary Markers: There are no explicit boundary markers or sanitization steps mentioned when interpolating the contents of these external files into the agent's context.
- Sanitization: The skill lacks mechanisms to validate the integrity or intent of the instructions contained within these repository-local files, which could lead to an attacker influencing the agent's output by modifying the configuration.
Audit Metadata