Flags and evals

Command	Purpose
vatio flags	Every open flag as markdown: the conversation, each tool call of that turn with your backend's response, and what the reply should have been
vatio flags --save [PATH]	The same, written to vatio-flags.md or PATH to hand to a coding agent
vatio flags --json	The open flags as JSON rows
vatio eval [--env NAME] [--agent SLUG]	Replay every eval case against an environment, wait for the judge, print the report; exits 1 if a case fails
vatio eval --no-wait	Start the runs and print their ids
vatio eval --show RUN_ID	One run's report

vatio eval defaults to the branch's environment on a branch and to preview elsewhere, and waits up to 900 seconds (--timeout N). A case that passes on live and fails on the environment you ran is listed first, as a regression.

Each flag with a note is an eval case. Replays call the environment's tools for real. The whole loop, and what resolves a flag, is in Improve your agent from flags.