If you have access to your document root, why go through a bot if you have access to the document files? It seems like it would be easier to just grep through all your html files in htdocs or wherever they are than to deal with error handling, following off site links, etc you might need to deal with a bot. Quicker too.