01 · Roasts
Star Vacuum
26 repos, 12 scored, and you've accumulated a grand total of 3 stars — SwarmSell's 2 stars are doing the heavy lifting. Your GitHub is a product launch factory where nobody RSVPs.
Hackathon Hopper
ReRoute (2 days old), SwarmSell (3 days), Zo (24 hours), FreezeFrame (3 weeks) — half your portfolio was born and abandoned inside a long weekend. Depth scores don't lie: burst commits ≠ shipped software.
Python Monoculture
89% Python by byte count. You've got Rust, Swift, and TypeScript repos, but your language entropy looks like a single crop farm. One blight and the whole portfolio wilts.
The License Nihilist
Checked your repos for licenses: found basically none. You're shipping multi-agent AI systems, iOS apps, and Rust CLIs into the void with zero legal clarity. OSS aspirations require more than vibes.
63 PRs, 6 Followers
You filed 63 PRs this year — more than most engineers — yet only 6 people follow you. Either you're PRing exclusively into your own repos or you're the most low-key contributor in UCSB history.
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% weight62C
- Depth15% weight65C
- Breadth10% weight65C
- Community10% weight40D
03 · Stats
365-day commit heatmap
89 active days
Language distribution
- Python89%
- JavaScript4%
- TypeScript2%
- Swift1%
- Cython1%
- C1%
- Other2%
04 · Numbers
Owned repos
non-fork
18
Commits
last 12 months
462
Followers
6
Joined GitHub
Jan 2020
05 · Top repos
adityasingh2400 /
SwarmSell
Hackathon-built multi-agent AI system for automating secondhand item listing across platforms. Uses Gemini video analysis, Browser-Use agents, and CDP screencasting to orchestrate concurrent marketplace submissions. Structured Python backend with tests and typed models, but unpolished and feature-incomplete.
adityasingh2400 /
ReRoute
Hackathon winner with 9 AI agents (Fetch.ai uAgents framework) competing across 5 resale routes. Typed Python/FastAPI backend, Gemini 3.1 integration, OpenCV media processing. Real application architecture with structured multi-file layout, but nascent (2 days old, no CI/tests, no external validation).
adityasingh2400 /
Recursor
Recursor is a well-crafted Rust utility integrating with Cursor AI agents for window focus management across macOS/Windows/Linux. Typed, documented, structured codebase with CI/CD but nascent adoption (0 stars, fresh repo from Feb 2026).
adityasingh2400 /
ivsunsets
Polished Next.js sunset forecast SPA for Isla Vista with TypeScript, Tailwind, Framer Motion, custom scoring algorithm, and interactive UI. Active but nascent (11 commits in 10 days), no tests/CI, no license.
adityasingh2400 /
Pepper
Niche iOS peptide tracking app with voice-first UX, 24-compound catalog, dosing math, and parser for stack imports. Typed Swift 6, tests included, but no CI/CD and no public adoption signals.
adityasingh2400 /
Zo
YC hackathon AI livestream-commerce project (24h build). Structured Python/TypeScript with typed core agents, comprehensive documentation (ARCHITECTURE.md, DESIGN_PRINCIPLES.md), CI/CD on GitHub Actions. Lacks public adoption but demonstrates substantial architectural scope (~1.2 MB codebase with multi-module agents, F
adityasingh2400 /
webbasketball
Web-based 3D basketball game with hand-tracking using MediaPipe, motion-controlled dribbling/shooting, and physics-based shot arc. TypeScript + React Three Fiber with functional architecture and unit tests, but pre-release stage with no stars/adoption yet.
adityasingh2400 /
ziri
Experimental distributed voice OS for Mac with multi-agent LLM orchestration, semantic memory, and always-on listening. Early-stage project (0 stars, 20 commits in ~3 weeks) with solid architectural foundation but incomplete feature set.
adityasingh2400 /
FreezeFrame
Ambitious multi-stage bullet-time sports replay system (4 phones → 3D moments) using Gemini generative fill + geometric reconstruction. Structured codebase (~2MB) with typed schemas, thoughtful architecture, but experimental—early stage (3 weeks old, 1 star), no tests/CI, no license.
adityasingh2400 /
website
Personal portfolio website in TypeScript using Next.js with gravitational orbit navigation and AI voice agent integration; typed and structured but lacks tests, CI, docs, and meaningful project documentation.
adityasingh2400 /
formfix
Basketball shot analyzer with AI-powered coaching via MediaPipe pose tracking. 6 commits in ~3 months with structured backend/frontend split, comprehensive docs (ARCHITECTURE.md, design.md, STATUS.md), but untyped Python and no tests/CI.
adityasingh2400 /
adityasingh2400
Personal profile repository with resume-style README and CI setup, but no substantive code artifacts. 30 commits over ~2.5 months, showing activity but minimal depth for a portfolio project lacking shipping examples.
06 · Timeline
- Jan 20, 2020Joined GitHub
- Dec 23, 2025Created formfix — AI-powered basketball shot form analyzer that uses pose and hand tracking to give instant mechanics feedback from video.
- Feb 2, 2026Created adityasingh2400
- Feb 2, 2026Created Recursor — The 'Bounce Back' utility for Cursor AI Agents — automatically saves and restores window focus
- Feb 2, 2026Created website — Innovative portfolio website with gravitational orbit navigation system
- Feb 27, 2026Created ziri — Distributed home voice OS with an always-on AI voice assistant for your Mac, smart home, music, and natural conversation.
- Mar 8, 2026Created ivsunsets — Predicts sunset quality for Isla Vista, CA — scored 0-100 with real-time weather data and animated sky visualizations
- Mar 21, 2026Created ReRoute — 🏆 First Place CSULB BeachHacks — Film one video of your unused stuff. Nine AI agents compete across five sale routes and execute the winning strategy automatically.
- Mar 25, 2026Created webbasketball
- Mar 28, 2026Created FreezeFrame
- Apr 4, 2026Created SwarmSell — Film your clutter. AI agents list it everywhere — simultaneously.
- Apr 16, 2026Created Pepper — peptide for all
- Apr 18, 2026Created Zo — YC Gemma 4 Voice Agents Hackathon Project
- Apr 24, 2026Most recent push to Zo
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.