01 · Roasts
Commit Heatmap? More Like Commit Dot
127 commits across a year and the heatmap looks like someone fired one pixel at a canvas. Weeks 1–30 are a complete void — 0 activity for half the year.
Serial Repo Abandoner
TaskAuth: 1 commit. green-commute-planner: 2 commits in 3 hours, README says 'under progress.' You don't start projects, you start READMEs.
Stars: 0/0/0/0/0/0/0/0/0/0/0/0/0/0
14 public repos, 0 total stars. Not a single one. The internet has issued a formal non-response to your entire portfolio.
PRs: Deeply Theoretical
0 external PRs, 0 issues opened, 98% solo development. You've built a WebRTC app, an AI code reviewer, and an SVM from scratch — and still haven't opened a single PR on anyone else's code.
License? Never Heard of Her
12 repos with real code, zero licenses. Your fraud detection pipeline, your PCA implementation, your peer-to-peer file transfer app — all legally ambiguous. MIT takes 30 seconds.
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% weight55D
- Quality20% weight62C
- Depth15% weight50D
- Breadth10% weight65C
- Community10% weight25F
03 · Stats
365-day commit heatmap
25 active days
Language distribution
- TypeScript41%
- Python38%
- HTML17%
- Jupyter Notebook3%
- CSS1%
- JavaScript0%
04 · Numbers
Owned repos
non-fork
14
Commits
last 12 months
127
Followers
2
Joined GitHub
Sep 2025
05 · Top repos
ryanmartin060708-code /
principal-component-analysis
Educational PCA implementation from scratch using NumPy. Well-structured project with comprehensive tests (38 passing), clear mathematical documentation, CLI interface, and visualization suite. Typed Python with proper project layout, examples, and solid engineering practices despite being a personal learning project.
ryanmartin060708-code /
pdfwizard
PDF Wizard: Personal image-to-PDF converter with full TypeScript, Tailwind UI, and pdf-lib backend. Polished indie project with README, structured components, and localStorage persistence, but nascent (18 days old, 0 stars).
ryanmartin060708-code /
support-vector-networks-reproduction
Educational SVM implementation reproducing Cortes & Vapnik (1995) from dual formulation. Well-structured with kernels.py, optimizer.py (SMO+QP), typed Python, and runnable experiments on synthetic/USPS/MNIST data. No tests, no CI, modest impact as personal research project.
ryanmartin060708-code /
qrcodegen
Personal QR code generator web app built with Next.js 15, TypeScript, and Tailwind CSS. Includes client-side generation, URL validation, dark mode, and deployment to Vercel. No tests or CI pipeline.
ryanmartin060708-code /
expensetracker101
Browser-based expense tracker built with Next.js 15, TypeScript, and Recharts. Personal project with polished UI, no production usage signals or external adoption. TypeScript + structured src/ with components, lib, types folders and meaningful README meet 50+ threshold; responsive dashboard with charts, filters, and CS
ryanmartin060708-code /
codesage-ai
Early-stage AI code review platform with Docker-based architecture, FastAPI backend, Next.js frontend, and RAG integration via Ollama and ChromaDB. Functional core with typed code and CI, but minimal project maturity (0 stars, 7 recent commits, created July 2024).
ryanmartin060708-code /
Disaster-dashboard
Portfolio civic tech dashboard for India disaster monitoring using vanilla HTML/CSS/JS, OpenWeather API, and Leaflet.js. Single-file monolith with live demo, 7 commits over 6 weeks, no tests/CI/license; meaningful alternate docs (DESIGN.md, ARCHITECTURE.md, STATUS.md) and structured planning offset missing type safety.
ryanmartin060708-code /
credit-fraud-detection
Personal ML portfolio project: end-to-end fraud detection with 6 models, Optuna tuning, SHAP explainability, and Streamlit dashboard. Well-structured Python codebase with typed code and modular src/ layout, meaningful documentation, but new repo (1 day old, 10 commits) with no tests, CI, or license.
ryanmartin060708-code /
peerdrop
TypeScript peer-to-peer file transfer app using WebRTC, React/Vite, Express, Socket.IO, and Prisma. Production-grade design (Apple HIG, SHA-256 verification, backpressure handling) but very recent (4 days old), 0 stars, no tests/CI, and only 5 commits.
ryanmartin060708-code /
green-commute-planner
Very early-stage TypeScript project (27 KB, 2 commits in 3 hours) with minimal README ("under progress"), no tests, CI, or license. Placeholder scaffold with gitignore present.
ryanmartin060708-code /
ryanmartin060708-code
Personal profile README with no functional code; empty scaffold created 2026-07-24 with only 4 commits in 14 minutes. Zero stars, no tests, no CI, no code files sampled.
ryanmartin060708-code /
TaskAuth
Empty scaffold with no files, commits, or documentation. Created and immediately abandoned with single initialization commit.
06 · Timeline
- Sep 11, 2025Joined GitHub
- May 21, 2026Created Disaster-dashboard — Real-time disaster monitoring dashboard for India with live weather (OpenWeather API), interactive flood zone maps (Leaflet.js), cyclone alerts, traffic conditions, and relief camp
- May 29, 2026Created support-vector-networks-reproduction — Reproduced the 1995 svm paper
- Jun 9, 2026Created principal-component-analysis — PCA From Scratch: Principal Component Analysis in Python
- Jun 11, 2026Created credit-fraud-detection — End-to-end credit card fraud detection platform — 6 ML models, Optuna tuning, SHAP explainability, and an interactive Streamlit dashboard with batch prediction and risk scoring.
- Jul 14, 2026Created codesage-ai
- Jul 19, 2026Created qrcodegen
- Jul 20, 2026Created expensetracker101
- Jul 22, 2026Created pdfwizard
- Jul 24, 2026Created ryanmartin060708-code
- Jul 26, 2026Created TaskAuth
- Jul 26, 2026Created green-commute-planner
- Aug 8, 2026Created peerdrop
- Aug 12, 2026Most recent push to peerdrop
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.