improve-codebase-architecture
Pass
Audited by Gen Agent Trust Hub on May 2, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill analyzes local codebase content and documentation to provide architectural insights. This behavior is transparent and aligned with its described functionality.\n- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it reads untrusted data from the repository (glossaries and ADRs) and uses it to construct prompts for sub-agents.\n
- Ingestion points: Processes files from the target repository, including CONTEXT.md and ADR files in the docs/adr/ directory.\n
- Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present when interpolating codebase content into sub-agent prompts.\n
- Capability inventory: The skill can modify local files (e.g., updating CONTEXT.md) and spawn recursive sub-agents via the Agent tool.\n
- Sanitization: No content validation or sanitization is performed on the ingested project data.
Audit Metadata