It reads your docs, tickets and code, uses your running app with the accounts you give it, and builds a map of what your product does and who can do what.
Cofactor only uses what you connect or hand it. Everything it learns goes into one map of your product.
Product docs, specs and PRDs tell Cofactor what your app is supposed to do.
Tickets show what changed recently and what your team is working on, so testing follows the product.
Reading the code helps Cofactor understand how features fit together and what a change might affect.
Cofactor signs in with the accounts and permissions you give it, so it learns what each kind of user can see and do.
Each run starts with what’s missing. That means areas it has never explored, areas it explored longest ago, and anything the last run flagged to come back to.
Each agent signs in as a persona with real credentials, so Cofactor learns which users can reach each part of your app.
The map is only updated at the end of a run, with what was confirmed. It never overwrites something a person has set.
Does it crawl my site?
No. Cofactor records the pages it visits while it runs flows, explores and tests, and builds the map from that. It doesn’t do a separate crawl of your site.
Does it read my documentation?
Yes. Connect your docs, tickets and code, and Cofactor learns from them alongside your running app. It only uses what you give it access to, and it doesn’t scrape anything.
What does covered mean here?
A capability is mapped when a flow is linked to it, and covered when that flow passed on its last run. Cofactor reports these separately because they mean different things.
Can I correct it?
Yes. You can edit any field on the map, and Cofactor will never overwrite a change a person made.
Join the waitlist, or book a call to walk through your surfaces.