improve-codebase-architecture
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands to inspect the repository (e.g.,
git log --oneline) and uses platform-specific commands such asxdg-open,open, orstartto display the generated HTML report to the user. - [EXTERNAL_DOWNLOADS]: The generated HTML report references styling and diagramming resources from well-known content delivery networks, specifically Tailwind CSS and the Mermaid.js library via JSDelivr.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the local repository that could contain malicious instructions.
- Ingestion points: Repository source files, git commit history, and issue text.
- Boundary markers: The instructions explicitly advise the agent to escape repository and issue text before embedding it into the HTML report.
- Capability inventory: The skill can write files to the system temporary directory and execute local commands for analysis and report viewing.
- Sanitization: The skill includes explicit instructions for data escaping and configures the Mermaid diagramming tool with a strict security level to mitigate potential cross-site scripting (XSS) or injection risks.
Audit Metadata