improve-codebase-architecture
Pass
Audited by Gen Agent Trust Hub on Jul 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill generates an architecture report and executes system-level commands (
xdg-open,open, orstart) to automatically open the file in the user's web browser. This execution path is determined dynamically using environment variables like $TMPDIR.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection vulnerabilities.\n - Ingestion points: The agent reads arbitrary source code, documentation, and commit history from the user's project directory (SKILL.md).\n
- Boundary markers: No specific boundary markers or instructions are provided to isolate the agent from instructions embedded within the project files.\n
- Capability inventory: The skill has the ability to write to the filesystem and launch external applications (browser).\n
- Sanitization: The provided HTML template (HTML-REPORT.md) initializes the Mermaid.js library with the
securityLevelset to"loose". This configuration disables built-in sanitization, potentially allowing malicious content found in the codebase to perform Cross-Site Scripting (XSS) when the report is viewed.\n- [EXTERNAL_DOWNLOADS]: The generated HTML report fetches the Tailwind CSS and Mermaid.js libraries from well-known content delivery networks (cdn.tailwindcss.com and cdn.jsdelivr.net).
Audit Metadata