Select Documentation
This step is where you give the agent the context it needs to interpret your protocol correctly. AuditAgent uses your README, design notes, and any extra context you provide to disambiguate intent and reduce false positives.
The step is optional, but the scan is noticeably better when you give the agent context. See How It Works for why context matters.
What you can attach
Four kinds of input are accepted.
Repository READMEs. Browse the list of README files detected in your repository and pick the ones in scope. Use the search box to filter.
Public URLs. Click + to add a URL field, paste any HTTP link, and add as many as you need. The agent fetches the page content. It does not follow nested links.
File uploads. Drag and drop PDF, MD, or MDX files from your computer. Useful for design docs, threat models, and external specifications.
Free-form notes. Anything else you want the agent to know. Paste it into the Additional Documentation field.
Currently, only the URL passed is fetched. We do not extract nested links yet.
When you are done, click Next.
Stay within the per-scan token count and file limit shown on the Selection Overview panel on the right of the screen.