The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The generated HTML report fetches styling and diagram scripts from well-known and established public CDNs (cdn.tailwindcss.com and cdn.jsdelivr.net). This is standard behavior for document rendering and does not represent a security risk.
[COMMAND_EXECUTION]: The skill uses platform-appropriate system commands (xdg-open, open, or start) to display the generated architectural report in the user's default browser. This is an intended user-experience feature for report visualization.
[PROMPT_INJECTION]: The skill processes project-specific data (glossaries, ADRs, and source code) to inform its architectural review, creating an indirect prompt injection surface.
Ingestion points: Reads CONTEXT.md, architecture decision records in docs/adr/, and explores the project via a subagent.
Boundary markers: No explicit markers are used to differentiate project data from agent instructions.
Capability inventory: The skill can write files to the local temporary directory and execute system commands to open the web browser.
Sanitization: There is no explicit sanitization described for the codebase content before it is incorporated into the agent's analysis or the HTML output.

improve-codebase-architecture