01 · Roasts
4-Minute Architect
tic_tac went from 'let me design an economy system with AI matchmaking' to final push in 4 minutes flat. Either you type at 900 WPM or you discovered the GitHub upload button.
README-Driven Development
user_count promises millions of users, PostgreSQL, Redis, and Docker in its README — the actual code is a 12 KB scaffold created and abandoned on the same afternoon. Bold claims, single afternoon.
The Incomplete Submission
VirusTotal-Telegram-Bot's bot.py is cut off mid-function at line ~160. You pushed a half-written file, closed the laptop, and called it a repo. At least the README is done?
Sprint Season
8 of 11 repos were created and last-pushed on the same day. Your git log looks less like a portfolio and more like a 72-hour hackathon with no judging panel.
The Typo That Ships
local_AI_agent's requirements lists 'panda>=0.3.1' alongside 'pandas'. Neither installs correctly together. The LLM integration works, but the dependency file doesn't.
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% weight57D
- Depth15% weight35F
- Breadth10% weight55D
- Community10% weight25F
03 · Stats
365-day commit heatmap
11 active days
Language distribution
- Python77%
- JavaScript14%
- CSS5%
- HTML3%
- Makefile0%
- Dockerfile0%
- Other1%
04 · Numbers
Owned repos
non-fork
11
Commits
last 12 months
56
Followers
1
Joined GitHub
Jun 2025
05 · Top repos
dk3yyyy /
football_predictor
Early-stage ML pipeline for football prediction with XGBoost, FastAPI, and Streamlit dashboard. Well-documented and structured with CI/tests, but nascent (12 commits in 2 days), zero adoption (0 stars), and untested in production.
dk3yyyy /
combosorter
Personal Python streaming utility for combo list processing with 18KB codebase, environment config, and 5 commits over 2.5 months. No tests, CI, or license; incomplete README; untyped Python without type hints.
dk3yyyy /
password-generator
Experimental password generator CLI+web tool with 2 commits in <24 hours. Has README, tests, CI, and structured layout but thin real-world adoption. Demonstrates working tooling with good features (entropy, passphrase, pattern detection) but minimal sustained development.
dk3yyyy /
dnslookup
Early-stage DNS CLI tool with async resolution, Rich formatting, and WHOIS integration. Clean architecture but just created (1 day old), no tests/CI, untyped Python, and zero adoption signals.
dk3yyyy /
tic_tac
Fresh multiplayer Tic-Tac-Toe game with economy system, AI, and Socket.io backend; untyped JS, no tests/CI, minimal git history (3 commits in 4 minutes), but functional full-stack architecture.
dk3yyyy /
sol-eth-wallet-scanner
Experimental Telegram bot for Solana/Ethereum wallet analysis. 15KB codebase with async architecture, typed code (main.py, services.py), and README docs. Only 1 commit in 10 months indicates initial dump with minimal sustained work.
dk3yyyy /
combo-splitter
Single-file Python utility for splitting large text files into chunks. Functional script with MIT license and README, but untyped, no tests/CI, minimal commits, and limited architectural scope—a practical one-off tool rather than sustained project.
dk3yyyy /
local_AI_agent
Early-stage Python AI agent for restaurant review analysis using LangChain + Ollama. Minimal scope (2 files), no tests, untyped, documented README but loose structure. Single-day sprint with 5 commits.
dk3yyyy /
user_count
Telegram bot scaffold with well-written README describing ambitious scalability claims (millions of users, PostgreSQL, Redis, Docker) but zero production signals. Untyped Python, no tests/CI/license, created and pushed same day. Tutorial-stage project.
dk3yyyy /
VirusTotal-Telegram-Bot
Fresh Pyrogram-based Telegram bot for VirusTotal scanning with 2GB file support. Created and pushed within hours (2026-05-06), minimal commits, incomplete bot.py cutoff mid-implementation, no tests/CI/license, untyped Python.
dk3yyyy /
dk3yyyy
Personal portfolio README only: 9 KB, no code, no structure. Owner intro with tech badges and contact links; neither a project nor a working artifact.
06 · Timeline
- Jun 24, 2025Joined GitHub
- Jul 22, 2025Created sol-eth-wallet-scanner — A professional, feature-rich Telegram bot for analyzing Solana and Ethereum wallet addresses. Get comprehensive portfolio insights with real-time market data, interactive navigatio
- Jul 24, 2025Created combo-splitter — A high-performance Python tool for efficiently splitting large text files into smaller, manageable parts. Perfect for processing large datasets, combo lists, wordlists, or any larg
- Feb 20, 2026Created combosorter — A high-performance, stream-oriented Python toolkit for processing and transforming massive combo lists. Features memory-efficient multi-gigabyte parsing, robust email:password extr
- Apr 13, 2026Created user_count — A scalable Telegram bot that tracks user registrations and notifies an admin of milestone achievements.
- Apr 20, 2026Created dnslookup — Professional DNS query CLI with Rich terminal output, asynchronous resolution, and integrated WHOIS lookups.
- May 4, 2026Created dk3yyyy
- May 5, 2026Created football_predictor — An end-to-end Machine Learning pipeline for football match predictions. Features automated data scraping, XGBoost forecasting, a FastAPI backend, and an interactive Streamlit dashb
- May 6, 2026Created VirusTotal-Telegram-Bot — A high-performance Telegram bot powered by Pyrogram (MTProto) and the VirusTotal v3 API. Features 2GB file scanning, interactive inline reports, and local caching.
- May 6, 2026Created tic_tac — A premium multiplayer Tic-Tac-Toe game with economy and AI.
- May 6, 2026Created local_AI_agent — A Python-based AI agent that processes and analyzes restaurant reviews using vector embeddings and local language models.
- May 8, 2026Created password-generator — A secure, feature-rich password generator with CLI and beautiful web UI
- May 8, 2026Most recent push to password-generator
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.