01 · Roasts
Serial Commitment-Phobe
9 repos analyzed, zero have tests. Not one. The neural-chess-bot has a TensorFlow CNN pipeline but apparently 'does it work consistently?' is a question for another lifetime.
The 48-Hour Architect
jaipur-wealth-index has 4 pipeline phases, H3 hexagons, K-Means clustering, and… 3 commits in 48 hours. You designed a skyscraper and then left after pouring the foundation.
Security Advisory Pending
Chat-System ships with SECRET_KEY='secret' hardcoded in auth.py. Line 4. On a system handling authentication. This isn't a bug, it's a philosophy.
Heatmap? More Like Heat-Dot
39 commits in a year, 50 out of 52 heatmap weeks are completely empty. Your entire GitHub contribution history fits in a long weekend.
Quantity Without Momentum
26 public repos, 0 total stars, 0 forks. You're shipping in a sealed room — lots of doors built, none of them open to the outside.
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
- Impact25% weight48D
- Consistency20% weight55D
- Quality20% weight36F
- Depth15% weight50D
- Breadth10% weight65C
- Community10% weight25F
03 · Stats
365-day commit heatmap
19 active days
Language distribution
- HTML33%
- Python24%
- JavaScript23%
- CSS13%
- TeX7%
- PLpgSQL1%
04 · Numbers
Owned repos
non-fork
9
Commits
last 12 months
39
Followers
3
Joined GitHub
May 2025
05 · Top repos
utkarshtyagi-253 /
jaipur-wealth-index
Experimental geospatial socioeconomic indexing pipeline for Jaipur using H3 hexagons, multi-API enrichment (Places, TomTom), and K-Means clustering. Modular 4-phase design with typed Python and structured architecture, but nascent (3 commits, <48 hrs old, no tests/CI/license).
utkarshtyagi-253 /
neural-chess-bot
Full-stack chess bot with CNN backend and Node.js/Express frontend. Python codebase lacks tests, CI, type hints, and license; untyped, modestly documented, with 6 commits in ~4.5 months suggests initial burst development.
utkarshtyagi-253 /
FIR-Report
Experimental FIR management system integrating Gemini API for legal brief generation. Typed-lang absent, no tests/CI/license, modest scope with 24 KB codebase and limited commit history (4 of last 30).
utkarshtyagi-253 /
noahweidig.github.io
Personal portfolio website (84MB) with README linking to live projects, CI enabled, MIT license, but untyped markup/static HTML, no tests, minimal source code samples visible, and created/pushed same day indicating initial setup.
utkarshtyagi-253 /
Chat-System
Small personal chat project (9 KB, 0 stars) with core real-time messaging via FastAPI WebSockets and MongoDB. Lacks tests, CI, type hints, and has hardcoded secrets and duplicate send logic.
utkarshtyagi-253 /
Electricity-price-predictor
A personal ML project deploying an XGBoost electricity cost predictor via Streamlit. Contains working code and a live demo, but lacks tests, CI, typing, license, .gitignore, and meaningful documentation of the model/data pipeline.
utkarshtyagi-253 /
utkarshtyagi-253
Personal portfolio/resume repo with styled README showcasing skills and interests. No actual code artifacts, just profile documentation. 14 commits over 12 days with CI configured but no tests or meaningful project delivery.
utkarshtyagi-253 /
Lecture-Hall-booking-system
Hackathon prototype for lecture hall booking with HTML/CSS/JS frontend, minimal architectural scope (8 files, 14 KB), no tests, CI, or license. Client-side only; demonstrates concept but lacks production readiness.
utkarshtyagi-253 /
utkarshtyagi-253.github.io
Personal portfolio website built with vanilla HTML/CSS/JS; minimal scope (17 KB), 2 commits in ~23 minutes, no tests/CI/license, documented only in README describing features rather than codebase.
06 · Timeline
- May 31, 2025Joined GitHub
- Jul 30, 2025Created Electricity-price-predictor — In this project I have created a machine learning model using Xgboost where I am predicting the cost in production of electricity and i have deployed this using Streamlit
- Sep 5, 2025Created Lecture-Hall-booking-system — in this repository i have created a system where admin can update the events and user can access the event schedule and user can access it by creating an account
- Mar 14, 2026Created neural-chess-bot — A deep learning chess bot that 'sees' the board using CNNs. Full-stack implementation using Python, Node.js, and Keras.
- Mar 22, 2026Created Chat-System — A scalable real-time chat backend and frontend built using FastAPI, WebSockets, and MongoDB, featuring live messaging, typing indicators, and online user tracking.
- Jun 1, 2026Created FIR-Report — Built an AI-powered FIR Report Management System that uses the Gemini API to generate structured FIR descriptions from user inputs.
- Jun 11, 2026Created noahweidig.github.io — Noah's Personal Website
- Jul 31, 2026Created utkarshtyagi-253
- Jul 31, 2026Created utkarshtyagi-253.github.io
- Aug 11, 2026Created jaipur-wealth-index — Geospatial Socioeconomic Index pipeline for Jaipur using H3 hexagonal grids, proxy data (DLC rates, POI density, traffic, demographics), Gemini 1.5 Flash OCR, and an interactive we
- Aug 12, 2026Most recent push to utkarshtyagi-253
07 · Compare
08 · Rubric
How this score was produced
Overall = Σ (category × weight) + gentle top-end curve
Tier thresholds
▸ How the pipeline works
- 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.
- 02Triage.A small model reads every repo's file tree + README and picks the 20 files per repo that actually reveal how you code.
- 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.
- 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.
- 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.