▸ This tool was built by an AI agent from Zoral
← RATE MY GITHUB

#547 — Top 69.0%

Zhreyu

Shreyas S

D

README enthusiast

Overall

0.0

/ 100

01 · Roasts

Notebook Hoarder Supreme

87% of your GitHub byte-count is Jupyter Notebooks. That's not a portfolio — that's a lab notebook collection. Real engineers ship .py files; you ship .ipynb files with 'Cell 1: import torch'.

Burst-and-Ghost Architect

quid: 6 commits in 2 weeks. Monoid: ~25 commits in 2 months. kamikaze: exists. You build entire systems in sprints, drop a technical_report.md, and vanish. GSoC taught you to ship — nobody taught you to maintain.

Tests Are For Other People

4 of your 6 analyzed repos have HAS_TESTS=no. Monoid is the lone hero with pytest. The other projects are just vibes plus TypeScript types, which is not the same thing.

47 Stars Across 45 Repos

That's a 1.04 stars-per-repo average. You've built CMS starters, AI research tools, knowledge graphs, and underground techno websites — and the internet has collectively given you 47 stars. The ideas are there; the marketing is not.

cms-testing: The 8-Second Repo

Created 2026-04-23, only commit pushed 8 seconds later. That's not a project — that's a `npx create-next-app` with a git push. It has a README explaining the project structure of a project that doesn't exist yet.

Built using

Zoral

Shadows one worker for a week, then takes over their job with zero extra setup. Behaves exactly like the original.

zoral.ai

02 · Category breakdown

  • Impact
    25% weight
    48D
  • Consistency
    20% weight
    60C
  • Quality
    20% weight
    57D
  • Depth
    15% weight
    55D
  • Breadth
    10% weight
    55D
  • Community
    10% weight
    40D

03 · Stats

365-day commit heatmap

216 active days

Less
More

Language distribution

7 langs
  • Jupyter Notebook87%
  • HTML7%
  • Python3%
  • TypeScript2%
  • CSS0%
  • Java0%
  • Other1%

04 · Numbers

Owned repos

non-fork

30

Commits

last 12 months

201

Followers

59

Joined GitHub

Jan 2022

05 · Top repos

Zhreyu /

kamikaze

45/100

Underground techno event website built with Next.js + TypeScript, featuring glitch UI effects, audio visualization, and Supabase serverless functions. Typed, structured layout with CI/CD, but no README and experimental scope.

I25Q60D50
CITyped
TypeScript★ 07mo ago

Zhreyu /

Monoid

45/100

CLI-first personal knowledge management system with AI-augmented notes, semantic search, and graph visualization. Typed Python codebase with comprehensive docs, tests, and CI, but experimental single-author project with zero adoption signals.

I25Q60D50
READMETestsCI
Python★ 09mo ago

Zhreyu /

quid

40/100

Early-stage research implementation of QUID (query expansion via masked diffusion). Demonstrates novel idea with substantial evaluation (1,321+ queries across BEIR benchmarks) and detailed technical documentation, but code is incomplete (no tests, CI), untyped Python, and git shows only 6 commits in ~2 weeks with no ac

I25Q45D50
README
Python★ 07mo ago

Zhreyu /

pr-sentinel

37/100

Early-stage AI-powered PR analysis platform using TypeScript, PostgreSQL, and Claude. WIP status, zero stars, typed + documented with structured monorepo architecture but no tests or CI configured.

I25Q52D35
READMETyped
TypeScript★ 08mo ago

Zhreyu /

zhreyu.github.io

22/100

Personal portfolio website with no external adoption, documentation, or project infrastructure. Bare HTML with moderate file size but no evidence of architectural depth or quality practices.

I5Q25D35
HTML★ 08mo ago

Zhreyu /

cms-testing

15/100

Next.js CMS starter template with minimal setup: typed TypeScript, README with project structure, but only 1 commit in ~8 seconds, no tests, no CI, and 25 KB codebase representing a bootstrap scaffold.

I5Q40D5
READMETyped
TypeScript★ 07mo ago

06 · Timeline

  1. Jan 2, 2022
    Joined GitHub
  2. Feb 23, 2025
    Created zhreyu.github.io — A page about myself
  3. Dec 28, 2025
    Created Monoid — Personal Knowledge Substrate
  4. Mar 6, 2026
    Created pr-sentinel
  5. Mar 20, 2026
    Created kamikaze — Website for da dawgs
  6. Mar 22, 2026
    Created quid — Queries Unmasked by Iterative Diffusion
  7. Apr 23, 2026
    Created cms-testing
  8. Apr 23, 2026
    Most recent push to cms-testing

07 · Compare

github.com/
Zhreyu · 6dmedian coder

08 · Rubric

How this score was produced

Overall = Σ (category × weight) + gentle top-end curve

CategoryWeightScoreContrib.
Raw total53.1
Top-end curve+3.4
Final overall56.5

Tier thresholds

S90–100Mass-producing humansA80–89Ship machineB70–79Solid engineerC60–69Getting thereD40–59README enthusiastF0–39GitHub tourist
▸ How the pipeline works
  1. 01Scrape.Pull every non-fork repo pushed in the last 90 days, plus your contribution calendar, followers, and language byte counts — straight from GitHub's REST & GraphQL APIs.
  2. 02Triage.A small model reads every repo's file tree + README and picks the 20 files per repo that actually reveal how you code.
  3. 03Grade each repo. All repos run in parallel through a fast scoring model that reads the picked files and rates each one independently on Impact, Quality, and Depth — with evidence citations.
  4. 04Aggregate. A larger reasoning model combines the per-repo scores with server-computed stats (heatmap, commit cadence, language entropy, follower count) to produce the 6-dimension profile score + roasts.
  5. 05Correct.Deterministic server-side checks enforce anchor-scale floors (e.g. a profile with 2,000+ public commits can't score 30 Consistency) and recompute the final verdict.

~90 seconds per profile, ~$0.25 in compute. Total of ~240 files read across your top-12 repos. One rating per GitHub account per day.

▸ Data sources & caveats
  • Heatmap & commit totals: GitHub GraphQL contributionsCollection — covers the last 365 days, includes private repos when the user has opted in (default).
  • Language %: byte totals across the top 30 owned non-fork repos.
  • Curve: a small upward nudge centered on raw score ≈ 70, capping at 100. Prevents specialists from being unfairly penalised for narrow breadth.
  • Anchor corrections: when server-measured signals (e.g. privateWorkLikely, multiRepoVolume, follower count) mandate a minimum category score, the aggregation step enforces it. These are signal-conditional, not identity-based floors.
Zhreyu · 56.5/100 — Rate My GitHub