01 · Roasts
Sprint God, Marathon Ghost
Living-Photos: 5-day hackathon, 122 tests, architecture docs, adapter factory, Inngest pipeline. spotify-for-hackers: 90 minutes, feature-complete, live on Vercel. You build fast — then vanish. The heatmap tells the whole story: ghost town for 25 weeks, then chaos.
97% Solo Act
soloPct=97%, totalPRsYear=0, totalIssuesYear=1 — you've opened exactly one issue on someone else's repo all year. With 85 public repos, you're clearly not shy about shipping, just… allergic to other people's code.
14 Stars Across 85 Repos
You've built a voice game, a 3D photo SaaS, an educational RAG dashboard, and an MCP gateway. The GitHub star market disagrees with your output rate: 14 total stars, 1 fork. That's 0.16 stars per repo. Marketing budget: $0.
License? Never Heard of Her
6 of 8 scored repos have no LICENSE file. Living-Photos went the other direction — Source-Available with commercial restrictions. So your default distribution policy is either 'take it, I don't care' or 'don't touch it ever'. No middle ground.
Python 81%, Python Repos: 0 Scored
81% of your language bytes are Python, yet every interesting project in your portfolio is TypeScript/JavaScript. Either your Python repos are locked in private purgatory or 67 repos of unscored Python are doing the heavy lifting nobody can see.
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% weight65C
- Quality20% weight72B
- Depth15% weight58D
- Breadth10% weight65C
- Community10% weight25F
03 · Stats
365-day commit heatmap
227 active days
Language distribution
- Python81%
- JavaScript9%
- C++4%
- Jupyter Notebook2%
- Cython1%
- TypeScript1%
- Other2%
04 · Numbers
Owned repos
non-fork
82
Commits
last 12 months
226
Followers
9
Joined GitHub
Feb 2020
05 · Top repos
anirxdh /
Living-Photos
Hackathon-winning fullstack Next.js/TypeScript app for transforming photos into walkable 3D memories using World Labs, FAL, ElevenLabs, and Stripe. Well-architected adapter pattern, typed, documented, and tested; source-available license limits adoption.
anirxdh /
spotify-for-hackers
Terminal-style Spotify UI with AI DJ voices (ElevenLabs) and LLM transitions (Groq). Next.js 16 + TypeScript, well-documented stack, but shipped in 90 minutes with zero tests/CI and no external adoption yet.
anirxdh /
Voice-companion
Voice-first AI companion built in TypeScript with Next.js, React, Framer Motion, and ElevenLabs TTS. Ambitious single-week project shipped with typed code, documented architecture, and structured multi-file layout including intent classification, orchestration, and animated colony UI.
anirxdh /
InsightEDU
InsightEDU is a React+Vite educational data dashboard with 8 page routes, RAG-powered AI chatbot, Pinecone/OpenAI integration, and responsive Plotly visualizations. Typed language missing; no tests/CI; ~27k codebase with decent structure but sparse docs.
anirxdh /
voices-of-the-last-world
Cinematic React/Vite web game featuring multi-agent debate with ElevenLabs voice synthesis. Polished UI, structured game loop, and sophisticated voice emotion tuning, but early-stage project with no tests or CI.
anirxdh /
YC-hack
TypeScript MCP orchestration gateway with React phone-UI widget. Early-stage (5 days old, 23 commits), no tests/CI/license, but typed + structured src/ layout + meaningful README describing architecture. Connects backend MCP servers, proxies tools, hosts local features (contacts, calling).
anirxdh /
CIVS
Early-stage Flask voting system using hand gesture recognition with TensorFlow CNN. Incomplete gesture detection code, inconsistent module structure (archived signdetect.py alongside active gesture.py), and no tests or CI pipeline. Shows architectural ambition but lacks polish.
anirxdh /
anirxdh
Personal GitHub profile repo with CV-style README and single snake animation workflow. No source code, just metadata and self-promotion. Minimal sustainable project content.
06 · Timeline
- Feb 1, 2020Joined GitHub
- Feb 6, 2025Created CIVS
- Jun 16, 2025Created anirxdh
- Jun 19, 2025Created InsightEDU
- Feb 21, 2026Created YC-hack — LARK - Multi-Agent Orchestration MCP
- Apr 23, 2026Created voices-of-the-last-world
- May 7, 2026Created spotify-for-hackers
- May 14, 2026Created Voice-companion
- May 16, 2026Created Living-Photos
- May 23, 2026Most recent push to anirxdh
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.