improve-codebase-architecture
Pass
Audited by Gen Agent Trust Hub on May 16, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by design, as it must ingest and analyze untrusted content from the user's codebase and documentation.
- Ingestion points: The skill reads
CONTEXT.md,docs/adr/, and various source code files across the repository using theExploresub-agent tool. - Boundary markers: No explicit delimiters or system instructions are defined to help the agent distinguish between its own operational instructions and data-driven content found in the codebase.
- Capability inventory: The skill is capable of spawning recursive sub-agents via the
Agenttool to perform exploration and design tasks, and it can modify documentation files likeCONTEXT.mdor create new ADR files based on its analysis. - Sanitization: There is no evidence of sanitization or filtering of codebase content before it is processed by the agent or used to influence the output of sub-agents.
Audit Metadata