improve-codebase-architecture
Pass
Audited by Gen Agent Trust Hub on May 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's primary purpose is architectural analysis and documentation. It guides the AI to read source code, suggest refactorings, and record architectural decisions (ADRs) or domain context.
- [EXTERNAL_DOWNLOADS]: The skill contains a reference to a GitHub repository (mattpocock/skills) for attribution and license purposes. This is a static documentation link and does not trigger automated downloads or remote code execution.
- [COMMAND_EXECUTION]: There are no shell commands or scripts included that would execute arbitrary code on the host system. The mention of an 'Explore' sub-agent refers to a high-level platform capability for codebase navigation.
- [DATA_EXFILTRATION]: While the skill is designed to read and analyze project files, there are no instructions or capabilities that would allow it to transmit this data to an external server or non-whitelisted domain.
- [PROMPT_INJECTION]: The instructions are clearly focused on technical refactoring tasks and do not attempt to bypass safety guidelines, reveal system prompts, or override agent constraints.
Audit Metadata