cdad roadmap

A check report tells you how illegible the repo is. It doesn't tell you where to start fixing that, and with a repo of any real size, "start everywhere" isn't a plan. cdad roadmap is the command that turns a score into an order of operations.

It reads cdad-report.md, the file cdad check just wrote, and asks you two more things for each capability the check step covered: how critical is this if it breaks, and how often does it actually get touched. Those two answers get multiplied against a third axis the check step already produced, the tribal knowledge density implied by that capability's legibility score. Criticality times frequency times tribal knowledge density gives you a ranking, not a vibe. The capability at the top of the list is the one where getting it wrong costs the most and a contract would help the most often.

The output is cdad-roadmap.md, written next to the report it came from. If you'd rather consume the ranking programmatically, --format json writes it as structured data instead of prose. --input <path> and--output <path> exist for the same reason the check command has them, so the two files don't have to live at the default location if your repo's conventions say otherwise.

Whatever sits at the top of the roadmap is what you hand tocdad init next. That's the whole point of ranking it in the first place, so the first capability anyone contracts is the one that was actually worth doing first.

cdad roadmap
cdad roadmap --input cdad-report.md --format json