01 · Roasts
The 6-Minute EDA Special
Student-Dataset-EDA was created and last pushed within 6 minutes of each other. That's not a project, that's a file upload with extra steps.
53 Stars, 0 Tests
EDAgent somehow pulled 53 stars with 8 commits, no type hints, no CI, and a 48 KB codebase. The GitHub algorithm is more impressed with your work than you are.
Testing? Never Heard of Her
Across 10 repos — ML pipelines, RAG agents, traffic analytics — HAS_TESTS=yes appears exactly zero times. Even your chess game ships untested.
Heatmap of Legends
Your public contribution heatmap is 38 consecutive weeks of void followed by a sudden burst of commits. GitHub's contribution graph looks like a starfield from a broken telescope.
License Collector (None Found)
Out of 10 repos with code, exactly 1 has a license. You're building on open-source giants and giving nothing back to the legal commons.
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% weight62C
- Consistency20% weight60C
- Quality20% weight57D
- Depth15% weight55D
- Breadth10% weight65C
- Community10% weight25F
03 · Stats
365-day commit heatmap
26 active days
Language distribution
- Jupyter Notebook40%
- Python33%
- JavaScript14%
- CSS11%
- HTML2%
- Dockerfile0%
04 · Numbers
Owned repos
non-fork
12
Commits
last 12 months
141
Followers
6
Joined GitHub
May 2022
05 · Top repos
Varn1t /
LoreLoop
Well-architected self-correcting RAG system with dual-graded verification and conversational memory, built in typed Python with structured multi-file layout, meaningful README, and ~1.2k kb codebase spanning agent orchestration, Streamlit UI, and CLI interface across 1 month of active development.
Varn1t /
TraffiQ
YOLOv8 + ByteTrack traffic analytics with Flask dashboard; typed Python, multi-file architecture, incident/speed detection; no tests/CI but 3-month development, ~1000 LOC, structured codebase.
Varn1t /
Varn1t.github.io
Personal portfolio site built in React + Vite showcasing AI/ML projects. Untyped JavaScript, no README or CI/tests, but well-architected with ~6260 KB codebase, structured components, and 30 recent commits across 3 months showing sustained effort.
Varn1t /
Varn1t
GitHub profile config repository with README showcasing 4 featured ML/MLOps projects. No source code files sampled; HAS_CI=yes but no tests, no license, no typed language. 323 KB with mixed commit activity suggests metadata/doc-only repo.
Varn1t /
Car-Price-Predictor
Early-stage personal ML project: Flask + scikit-learn car price predictor with 162 KB codebase, 17 commits over ~6 weeks, typed backend code missing, no tests/CI. Documented README but sparse architecture.
Varn1t /
SkyFlow-MLOps
Early-stage MLOps pipeline for rain forecasting using Airflow, MLflow, and LLM-gated deployment. Well-architected with containerization and modern ML stack, but < 48 hours old with only 2 commits; untyped Python, no tests, no CI, no license.
Varn1t /
Chess_game
Personal chess game project in Python with minimal documentation, no tests/CI, and recent activity (last push Feb 2026) but sparse commits (7 of last 30) over ~2.5 years of existence.
Varn1t /
EDAgent
An agentic EDA system using LangGraph and Ollama with 9-stage pipeline. Functional but bare-bones: no tests, no CI, no type hints, no license, minimal commits (8 of last 30), and only ~2 weeks old with 48 KB codebase.
Varn1t /
Criminal-Record
A single-developer crime record CLI tool built over ~3 weeks with PostgreSQL backend. Untyped Python, no tests or CI, basic structure with .env config. One-shot educational project with minimal adoption signals.
Varn1t /
Student-Dataset-EDA
Jupyter-based student performance EDA with 1,000-record dataset analysis. Very recent repo (created 2026-04-11, last push same day), minimal commit history, single notebook analysis.
06 · Timeline
- May 25, 2022Joined GitHub
- May 25, 2022Created Varn1t — Config files for my GitHub profile.
- Oct 3, 2023Created Chess_game
- Feb 27, 2026Created TraffiQ — AI-powered traffic intelligence platform for real-time vehicle analytics, congestion prediction, and smart urban mobility systems.
- Feb 28, 2026Created Varn1t.github.io — Personal portfolio showcasing AI systems, MLOps, agentic workflows, and intelligent applications.
- Feb 28, 2026Created Car-Price-Predictor
- Apr 11, 2026Created Student-Dataset-EDA — Exploratory Data Analysis on a 1000-student dataset examining how demographics and socioeconomic factors influence Math, Reading, and Writing scores.
- May 1, 2026Created LoreLoop — An agentic, self-correcting RAG system powered by LangGraph, FAISS, and Ollama that verifies groundedness and relevance through dual-graded answer validation, conversational memory
- May 2, 2026Created Criminal-Record
- May 10, 2026Created EDAgent — Multi-agent exploratory data analysis system with autonomous insights, visualization, preprocessing, and reporting workflows.
- May 24, 2026Created SkyFlow-MLOps — An autonomous, agent-governed MLOps platform that orchestrates data ingestion, model training, experiment tracking, and deployment decisions using Airflow, MLflow, LangGraph, Ollam
- May 26, 2026Most recent push to Varn1t
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.