01 · Roasts
Sprint King, Maintenance Stranger
Attention-Is-All-You-Need: 14 minutes. Stock-Trend-Predictor: 40 minutes. Next-Word-Predictor: 2 hours. Trolley-Problem-Simulator: 72 hours. You code like you're speed-running a hackathon and then immediately lose interest.
0 Tests, 8 Repos, No Regrets
Not a single HAS_TESTS=yes across all 8 analyzed repos. Eight projects. Zero test files. You've built a mock interview app, a trolley problem simulator, and an art studio — and apparently test-driven development is the real trolley problem you keep dodging.
The Jupyter Notebook Iceberg
71% of your GitHub is Jupyter Notebooks, yet your actual shipped work is all TypeScript Next.js apps. Your language distribution is having an identity crisis — pick a lane, or at least get the notebooks out of the driver's seat.
1 Star. 26 Repos.
One star. Earned on QuillKeys — probably from yourself. With GitSaga, NexterView, Mosaik, and a trolley problem simulator all live, you're building a portfolio faster than the internet can find it.
Joined December, Already Philosophizing
Four months on GitHub and you've already shipped an app about the trolley problem. Most people spend years on GitHub before existentially questioning whether to pull the lever. Respect the pace, question the test coverage.
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% weight56D
- Consistency20% weight55D
- Quality20% weight62C
- Depth15% weight50D
- Breadth10% weight60C
- Community10% weight25F
03 · Stats
365-day commit heatmap
72 active days
Language distribution
- Jupyter Notebook71%
- TypeScript24%
- JavaScript2%
- Python2%
- CSS1%
- HTML0%
04 · Numbers
Owned repos
non-fork
26
Commits
last 12 months
172
Followers
6
Joined GitHub
Dec 2024
05 · Top repos
bedigambar /
QuillKeys
TypeScript literary typing test with React 18, Zustand, Tailwind. Shipped with core features (WPM tracking, keyboard heatmap, history, themes) but lacks tests, CI, and documentation beyond README. Personal project showing structured execution.
bedigambar /
Mosaik
Early-stage Next.js dithering art studio with typed code, structured src/, and ambitious feature set (Floyd-Steinberg, ASCII, physics engine, video export). Personal project launched April 2026, 0 stars, minimal community adoption yet.
bedigambar /
Trolley-Problem-Simulator
A polished, interactive ethical dilemma simulator built with Next.js, React, and TypeScript that explores utilitarian vs. deontological reasoning through 8-12 trolley problem variants. Well-styled, feature-rich quiz interface with animations, timed mode, and aggregate statistics tracking.
bedigambar /
GitSaga
TypeScript Next.js app that transforms GitHub commit histories into AI-narrated fantasy stories via Groq LLM. Typed, documented with README, structured layout, SSR pages, OAuth auth, streaming UI. Created Feb 2026 with 12 of 30 recent commits.
bedigambar /
NexterView
Newly-launched AI mock interview platform built with Next.js 16, TypeScript, Prisma, and Gemini API. Typed, documented, and architecturally sound, but brand-new (created 2026-03-22), zero adoption signals, and lacks test coverage and CI/CD.
bedigambar /
Stock-Trend-Predictor-Model
Jupyter Notebook-based LSTM stock price predictor with PyTorch training script and Streamlit web app. Zero stars, 5 commits in ~40 minutes on 2026-02-07. Clean typed Python code and documentation present, but experimental scope without production adoption signals.
bedigambar /
Next-Word-Predictor-LSTM
Educational LSTM next-word predictor trained on Medium articles. Jupyter-based implementation with model architecture and training code but no tests, CI, or reproducible artifacts beyond notebook format.
bedigambar /
Attention-Is-All-You-Need
One-shot educational implementation of the Transformer architecture from scratch in PyTorch. Clean single-file code with comprehensive class comments but no tests, CI, typing hints, or multi-file structure. Created and last pushed same day (2026-03-25, 14 minute window).
06 · Timeline
- Dec 2, 2024Joined GitHub
- Oct 6, 2025Created QuillKeys — QuillKeys - Literary Typing Test
- Feb 5, 2026Created Next-Word-Predictor-LSTM — A deep learning project that uses Long Short-Term Memory (LSTM) neural networks to predict the next word in a sequence.
- Feb 7, 2026Created Stock-Trend-Predictor-Model — A LSTM-based machine learning project for predicting stock price trends using historical data from Yahoo Finance.
- Feb 20, 2026Created GitSaga — Turn your GitHub commit history into an epic AI-narrated story with GitSaga.
- Feb 28, 2026Created Trolley-Problem-Simulator — An interactive ethical dilemma simulator that explores moral philosophy through the lens of the classic trolley problem and its many variants.
- Mar 22, 2026Created NexterView — NexterView - AI-powered mock interview platform.
- Mar 25, 2026Created Attention-Is-All-You-Need — This repository provides a crystal-clear, scratch-built PyTorch implementation of the Transformer.
- Mar 27, 2026Created Mosaik — An advanced, interactive browser-based studio for creating stunning dithered art, retro ASCII animations, and custom dot matrix graphics.
- Apr 1, 2026Most recent push to Mosaik
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.