CodeOak tree logoCodeOak
DataAI Study BuddyAboutBlog
Join Now
Join Now

CodeOak AI · Strengthen Your Coding Roots.

TermsPrivacyCookiesSupport
← The Grove · all articles
#CodeOak · Update

What we shipped in May: signals over vanity metrics

A redesigned post-batch panel, skill coverage maps, and time-per-question as a first-class signal — the May roundup of everything that landed in Canopy.

BhanuFounderJune 1, 2026
2 min read
OAK

May was about one idea: a streak counter tells you that you showed up, not that you got better. So we spent the month replacing counters with signals — measurements that only move when your skill does — across Canopy, CodeOak's dashboard.

What landed:

  • A redesigned post-batch panel that explains the next step instead of just unlocking it.
  • Skill coverage maps — a topic isn't "done," it's covered to a level.
  • Time-per-question promoted to a first-class signal.

Readiness, not streaks

The post-batch panel got a full redesign. Instead of "12/15 solved," you now see which skills you actually proved, which are still thin, and what the system opened next — each with a one-line reason. The point is to answer the only question that matters after a batch: am I closer to ready, or just busier?

Skills you can see

CodeOak grades the output of your work, not your effort, so the dashboard should reflect real movement. Two changes make that visible:

  • Skill coverage maps replace binary "complete" flags. A topic moves from not started, to in progress, to covered — at a level — so you can see where depth is still missing.
  • Time-per-question is now shown alongside accuracy. Solving correctly but slowly is a different signal than solving fast, and the panel treats it that way instead of hiding it.

These are the same signals the roadmap reads from when it decides your next batch — so what you see in Canopy and what the system does next come from one source of truth, not two.

What's next

June is about history and recovery: a readiness view that tracks how your signals move over time, and per-skill batches that target the areas a coverage map flags as thin. Open your dashboard — your next batch already reflects the new signals.

For the full picture of how assessment, roadmap, and batches fit together, see CodeOak is live.

FAQ

What is Canopy? Canopy is CodeOak's dashboard. It surfaces the signals that matter — accuracy, time per question, skill coverage, and batch state — plus whether you've earned the next step, instead of vanity counters like total problems solved.

Why drop streaks and solved-counts? Because they measure attendance, not ability. A count goes up whether or not you improved. Signals like skill coverage and time-per-question only move when your actual solving does, which makes them honest inputs for the roadmap.

Do these signals change what I practice next? Yes. The roadmap reads the same skill-coverage and performance signals shown in Canopy to generate your next batch, so the dashboard and the system stay in sync.

#CodeOak#Update
Share
Written by
Bhanu

Founder of CodeOak. Building the system that turns 'I don't know what to practice' into a roadmap that decides for you. Panda profile — steady, methodical, table-first.

Keep reading

All articles →
OAK#CodeOak · Product

CodeOak is live: Python & SQL prep that builds your roadmap, not a question bank

Bhanu · 5 min
PY#Python · pandas

pandas groupby that finally clicks: split, apply, combine

Bhanu · 3 min
SQL#SQL

Window functions, finally explained without the jargon

Bhanu · 3 min