rightcode-improve-codebase-architecture
Pass
Audited by Gen Agent Trust Hub on May 13, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is primarily instructional and defines a clear methodology for codebase improvement based on established software engineering principles. It helps the agent reason about code depth, seams, and adapters to improve testability and maintainability.
- [COMMAND_EXECUTION]: The skill uses the
Agenttool for codebase exploration (withsubagent_type=Explore) and to generate parallel interface designs. These operations are essential for the skill's functionality and are constrained to the agent's operating environment for the purpose of analysis. - [PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it is designed to ingest and analyze external codebase content.
- Ingestion points: Project source code files,
CONTEXT.md, and Architecture Decision Records (ADRs). - Boundary markers: Absent; the skill does not explicitly use delimiters to separate source code data from its internal reasoning instructions.
- Capability inventory: The skill utilizes the
Agenttool for exploration and design generation, and can write updates to documentation files (CONTEXT.mdand ADRs). - Sanitization: Absent; the skill does not specify validation or filtering of the codebase content it processes. However, the risk is mitigated by a human-in-the-loop 'grilling' process where users review all proposed changes.
Audit Metadata