The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches resources from well-known services to render the architectural report.
Downloads Tailwind CSS from https://cdn.tailwindcss.com for layout styling.
Downloads the Mermaid diagramming library from https://cdn.jsdelivr.net/npm/mermaid via JSDelivr.
[COMMAND_EXECUTION]: Uses standard system commands to facilitate the architecture review process.
Executes git log --oneline to identify codebase hot spots.
Invokes system openers like xdg-open, open, or start to display the generated HTML report to the user.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data which could potentially influence its output or the generated report.
Ingestion points: Reads files from the local codebase, commit history via git log, CONTEXT.md glossary, and Architecture Decision Records (ADRs).
Boundary markers: None specified for the sub-agent exploration or the report generation phases.
Capability inventory: Writes HTML files to the system temporary directory and opens them in the default web browser.
Sanitization: No explicit sanitization or escaping is mentioned for content interpolated into the HTML report or Mermaid diagrams.
[DYNAMIC_EXECUTION]: Generates and executes web content locally.
Assembles an HTML file at runtime using Tailwind and Mermaid and opens it in a browser context.
Configures Mermaid with securityLevel: "loose" in HTML-REPORT.md, which disables some of the library's internal sandboxing and allows for more complex (and potentially risky) HTML/script rendering within diagrams.

improve-codebase-architecture