INTRO
I have a problem - I can't do things blindly without seeing progress. When I run, I need a tracker with pace and distance. At work - Jira, metrics, dashboards. Everything visible: what's done, what's in progress, where I'm stuck.
With LeetCode - none of that. You solve problems and all you get back is a counter.
- Am I ready for interviews or not?
- Is 150 problems a lot or not enough?
- Which topics are weak?
- What should I review?
LeetCode doesn't answer any of this. Just a damn solved counter.
That wasn't enough for me. Started thinking about how to fix it.
LISTS
It all starts with picking a list - Blind 75, Grind 75, NeetCode 150, Grind 169, Combined 212 (NeetCode 150 + Grind 169 merged). Pick one, everything revolves around it: progress, analytics, reviews. No need to think "which problem to solve" every time - the system suggests the next one from your list: new problems + ones due for review.
SELF-ASSESSMENT
What does "solved a problem" actually mean?
At work, a bug isn't just "exists/doesn't exist". It has severity, priority, affected users. Based on that, we decide - fix now or later.
Applied the same approach to problems. Not "solved/not solved", but assessment by criteria:
- Help: solo → hint → fully followed solution
- Readiness: would solve in interview → maybe → definitely not
- Understanding: can explain → remember steps → no idea why it works
This builds the real picture. A problem can be "solved" but rated as "needs redo".
Another thing - speed. Interview time is limited. If you spend an hour on a Medium at home - that's a problem.
LC has a timer, but you need to remember to start it. I kept forgetting.
So I did it differently: open a problem page → start screen appears → click "Solve" → timer starts automatically. Got Accepted or hit pause → timer stops. No manual actions.
TIERS
When problems piled up, looking at scores became pointless. A bunch of numbers, nothing stands out.
Borrowed an idea from gaming - tier lists. Simple: S is best, D is bottom. Five letters instead of endless numbers.
Now each problem has a tier. S - know it well. D - don't know it, need to redo. A, B, C - in between.
Open the list - immediately see the pattern. These topics are green, all good here. These are red - come back to these.
REVIEWS
Spaced repetition - an old idea. Anki, flashcards, languages - works everywhere. The better you know it, the less often you review. The worse - the more often.
Same for LC. D-tier problem - come back tomorrow. S-tier - in a month or two.
At the end of the survey, you choose when to review. System suggests an option, but you decide. Home shows what's due.
INTERVIEW READINESS SCORE
Okay, we have:
- Assessment for each problem (survey)
- Tiers (S/A/B/C/D)
- Solve time
- Review schedule
But the main question remains: "Am I ready for interviews?"
Not "how many problems solved". Not "what's my streak". Actually - ready or not?
Here's what's usually recommended:
Time: Easy ≤10min, Medium ≤20min, Hard ≤35min - that's normal. Longer - problem.
Categories: Arrays, Trees, Graphs - critical, asked everywhere. DP, Stack - important. Math, Bit - less critical.
Coverage: Solving 50 problems well > solving 150 poorly. But if you only solved 30% of the list - also a problem.
Combined all this into one number - Interview Readiness Score (0-100).
It's not a job offer guarantee - it's a mirror. Shows where you are based on your own data.
Components: solution quality, topic coverage, speed, self-assessment, consistency. Minus penalties for weak spots.
Penalties: Gap in Graphs penalizes harder (×1.5) than gap in Bit Manipulation (×0.5).
Levels: 80+ ready, 60-79 almost, 40-59 on track, below 40 needs work.
One number instead of chaos. Not a guarantee, but a compass.
Below on the same page - detailed breakdown by each component. Each bar is clickable: tap → see specific problems dragging you down or up.
THE TOOL
You could do all this in Excel. I tried - gave up after two weeks. Manual entry, switching between tabs, zero automation.
That's how the idea for a Chrome extension was born - the only way to integrate with LeetCode and work right on the page.
Everything automated: timer starts itself, survey pops up after Accepted, tiers are calculated, reviews are scheduled. All you do is solve problems and answer honestly.
Three main pages:
Home - progress visualization for your list, current streak (+ 5 freezes per month if you miss), "Start Solving" button (system picks the next problem: new ones + scheduled reviews).
Tasks - all problems in a list. Smart filters directly by score components: Quality, Time, Consistency. See what's dragging you down - fix it.
Analytics - Interview Readiness Score and detailed breakdown by each part of the formula.
What it does: popular lists, auto-capture, auto-timer, survey, tiers, reviews, Interview Readiness Score, smart filters, Start Solving.
What it doesn't do: doesn't solve for you, doesn't give hints.
I called it HashTry.
If you want to try it without signing up — DM me, I'll send test account details
Pricing: 14 days free. Then you buy time: $6 for one month, $11 for two, $15 for three. Not a subscription - buy and use. Runs out - read-only, your data stays.
Link: Chrome Web Store
Link: HashTry.io
TL;DR
- Chrome extension for LeetCode
- Pick a list (Blind 75, NeetCode 150, etc.) - everything revolves around it
- Start Solving - picks the next problem from your list
- Auto-timer - starts when you begin solving
- Survey after problem → tiers (S/A/B/C/D) and spaced repetition
- Interview Readiness Score (0-100) - ready for interviews or not
CLOSING
How do you currently track your progress? Excel, Notion, in your head, not at all?
If you try HashTry - let me know what you think. Bugs, ideas, criticism, roasts - all useful.







Top comments (0)