Skip to main content

I failed a milestone validation

A failed milestone is information, not a setback. I have a specific protocol for what happens next, and it depends on how many times you have missed this same milestone.

After the first failure​

I queue a retry_prep task on your roadmap for the next study day. The retry_prep is a focused review of the exact concepts you missed in the validation — not the entire milestone. It typically runs shorter than the original milestone work because the gaps are narrower.

Once the retry_prep is complete, you re-attempt the validation. Most milestones pass on the second try.

After the second (and beyond) failure​

If you fail the same milestone twice or more, repeating the same prep is the wrong answer. I run complete-milestone-ai to restructure the approach instead of repeating it. That can mean:

  • A different angle on the same concepts (concrete examples instead of abstract definitions, or vice versa)
  • Splitting the milestone into smaller pieces with their own validations
  • Pulling in a prerequisite topic I now believe is the real blocker
  • A different session format entirely — drill, dialogue, or applied scenario

The restructured plan replaces the existing one. You do not see the old prep tasks come back.

What you should not do​

  • Do not try to brute-force the validation by retrying it back-to-back. Each attempt costs credits and the third attempt with no remediation almost certainly fails the same way.
  • Do not skip the retry_prep. The validation is designed to test the gaps the retry_prep covers — going in cold defeats the system.

Why ARIA restructures instead of repeating​

Repeating something you have already failed is the worst signal a study system can send. It tells you "do the same thing again and hope". That is not what I do. The whole point of the milestone validation system is to catch a gap and respond intelligently to it.

For the full protocol — pass thresholds, validation format, what gets retried — see Milestone validation.