01 · Roasts
Sprint King, Ship Nothing
27 public repos, 0 stars, 0 forks. quant-statarb-lab was created and last-pushed 52 seconds apart. The velocity is impressive; the landing is nonexistent.
README-Driven Development
Every repo ships with ARCHITECTURE.md, DESIGN.md, STATUS.md, and PROJECT_STATE.md — sometimes before the actual code compiles. newrepo's server.ts literally cuts off mid-function, but the design doc is immaculate.
Commitment Issues
The heatmap is a ghost town for 46 of 52 weeks. Your entire visible GitHub career happened in the last 3 weeks of April 2026. GitHub thought you were a bot until yesterday.
Interview-Defensible™
quant-statarb-lab's README literally says 'interview-defensible' as a feature. Shipping code to impress interviewers instead of users is a valid strategy — just don't be surprised when your star count matches your follower count: zero.
Lone Wolf of the Heatmap
92% solo commits, 0 followers, 0 following. You've opened 11 PRs this year but follow zero people. Neil codes in a vacuum sealed with CUDA and self-doubt.
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% weight68C
- Consistency20% weight65C
- Quality20% weight72B
- Depth15% weight65C
- Breadth10% weight65C
- Community10% weight25F
03 · Stats
365-day commit heatmap
19 active days
Language distribution
- Python57%
- TypeScript17%
- HTML10%
- JavaScript6%
- C++6%
- Cuda2%
- Other2%
04 · Numbers
Owned repos
non-fork
20
Commits
last 12 months
131
Followers
0
Joined GitHub
May 2020
05 · Top repos
Neil-N4 /
6dof-rocket-trajectory-simulator
Well-engineered 6-DOF rocket simulator with typed Python, C++ backend, comprehensive docs (design.md, ARCHITECTURE.md), CI/CD pipeline, and validation suite. Personal portfolio project with non-trivial physics engine and dual-language architecture.
Neil-N4 /
reward-model-bias-auditor
Frontier evals framework for auditing reward-model robustness via semantic-preserving perturbations and evolutionary exploit search. Well-structured Python codebase with typed interfaces, comprehensive documentation, and multi-stage attack-defense pipeline.
Neil-N4 /
cuda-path-simulation
Educational CUDA+C++20 Monte Carlo pricing engine with CPU baseline, variance reduction, Greeks estimation, and comprehensive validation scripts; well-documented but nascent (6-day-old with 24/30 commits, 0 stars).
Neil-N4 /
patchflow
Early-stage CLI tool for branch scope analysis and cleaning. Typed Python with organized structure, tests, CI, and TUI. No license; very recent (created April 9, 2026, 30 commits in ~1 day). Pre-product-market validation phase.
Neil-N4 /
market-filter-cli
Fresh stock-screener CLI with clear multi-file architecture (fetcher, filters, formatters, reporter), typed Python, README, and test coverage. Created 2026-03-21 with 3 commits; demonstrates real problem-solving but early-stage personal project with no CI or license.
Neil-N4 /
formal-math-benchmark
Formal math benchmark research prototype measuring claim-level reasoning coverage in LLM olympiad-style math solutions via Python evaluation + Lean 4 proof obligations. Fresh prototype with typed code, design docs, tests, but minimal adoption and very recent creation (same-day push).
Neil-N4 /
stat-arb-pairs
Statistical arbitrage pairs trading engine with cointegration screening, Kalman filters, and walk-forward backtesting. Experimental 1-day-old project with typed Python, structured src/, and comprehensive README. No tests in suite yet despite docstring references.
Neil-N4 /
meet-website
Next.js student platform website (Meet!) showcasing a campus social/event discovery concept. Typed, structured multi-file layout with Tailwind + Framer Motion animations, but no tests/CI, zero stars, and created within the past 24 hours with single commit.
Neil-N4 /
yc-voice-agents-hackathon
Hackathon project with alternate docs (design.md, ARCHITECTURE.md, STATUS.md) and structured src/ but nascent—131 KB, 0 stars, 30 commits in ~6 hours, no tests/CI. Early-stage experimental work without product presence.
Neil-N4 /
quant-statarb-lab
Personal portfolio project: interview-defensible stat-arb framework with synthetic data, pair screening, walk-forward backtesting, and risk metrics. Typed Python, structured, documented, tested, but brand new (created 2026-04-01, 1 commit), zero adoption.
Neil-N4 /
newrepo
Hackathon scaffold for voice-to-multiapp translation built in 24 hours. TypeScript server, Python brain using Gemma/Whisper, but incomplete: server.ts cuts off mid-response, no working routes beyond API stubs, missing core STT integration, no tests or CI.
Neil-N4 /
Neil-N4
Personal GitHub profile repo with only a README containing bio and contact links. No code, no projects, no meaningful substance—purely a profile card scaffold.
06 · Timeline
- May 21, 2020Joined GitHub
- Mar 21, 2026Created market-filter-cli
- Mar 25, 2026Created 6dof-rocket-trajectory-simulator — 6-DOF rocket trajectory simulator with RK4 integration, staging, atmospheric drag, and Max-Q/apogee analysis
- Mar 25, 2026Created cuda-path-simulation — C++20/CUDA Monte Carlo path simulation with CPU baseline, Nsight profiling, and benchmark tooling
- Mar 26, 2026Created Neil-N4
- Mar 26, 2026Created formal-math-benchmark — Benchmark for separating math answer accuracy from proof-valid reasoning via claim-level evaluation and Lean 4 obligations.
- Mar 26, 2026Created reward-model-bias-auditor — Evals framework for auditing reward-model robustness under semantic-preserving rewrites and reward-hacking cues.
- Mar 30, 2026Created stat-arb-pairs — Statistical arbitrage pairs trading engine with cointegration screening, Kalman hedge ratios, and walk-forward backtesting.
- Apr 1, 2026Created quant-statarb-lab — Statistical arbitrage research framework with synthetic market generation, pair screening, walk-forward backtesting, and risk-aware spread trading.
- Apr 9, 2026Created patchflow — Local-first CLI that cleans messy branches into mergeable PRs and explains merge blockers.
- Apr 17, 2026Created meet-website — Website for Meet!, a student-focused platform for discovering people, communities, and campus events.
- Apr 18, 2026Created yc-voice-agents-hackathon
- Apr 19, 2026Created newrepo
- Apr 19, 2026Most recent push to newrepo
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.