01 · Roasts
137 Repos, 73 Commits
You have 137 public repos but only 73 commits in the past year. That's one commit every 5 repos. You're not building a portfolio — you're hoarding domain names.
81% Graveyard Rate
staleRepoRatio=0.81: four out of every five repos you've ever touched haven't seen a push in over two years. Your GitHub is less a portfolio and more an archaeological dig.
Jupyter Notebook: 66%
Two-thirds of your codebase is Jupyter Notebooks. That's not software engineering — that's a very elaborate note-taking habit with matplotlib imports.
0 PRs, 0 Issues, 94% Solo
totalPRsYear=0, totalIssuesYear=0, soloPct=94%. You've built a paragliding apparatus that only flies over your own backyard. Open source is a team sport.
Burst Builder Syndrome
bandhu: 7 days. agentgate: 17 days. fredis: 3 days. NatureFix: 25 days. You sprint like a pro and then vanish. The heatmap is 40 weeks of zeros punctuated by one very caffeinated month.
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% weight48D
- Consistency20% weight60C
- Quality20% weight72B
- Depth15% weight55D
- Breadth10% weight65C
- Community10% weight40D
03 · Stats
365-day commit heatmap
29 active days
Language distribution
- Jupyter Notebook66%
- TypeScript14%
- JavaScript8%
- Python5%
- CSS4%
- HTML2%
- Other1%
04 · Numbers
Owned repos
non-fork
52
Commits
last 12 months
73
Followers
29
Joined GitHub
Sep 2016
05 · Top repos
RaiVaibhav /
fredis
Instructional Redis implementation in TypeScript with 3 of 7 milestones complete. Strict types, comprehensive tests, structured docs (docs/ folder, design.md, ARCHITECTURE.md), and clean architecture, but nascent project (4 commits over 3 days) with zero adoption signals.
RaiVaibhav /
NatureFix
Polished Next.js product marketing site for a commercial mountain-experience business. Well-typed TypeScript, structured component architecture, rich documentation (README + design.md + ARCHITECTURE.md + STATUS.md), and thoughtful UX. No tests or CI. ~26 commits over ~25 days, 41MB codebase.
RaiVaibhav /
bandhu
Prototype mental-health companion app with real backend (FastAPI + Postgres/pgvector), frontend (React 19 + TypeScript), and 12-stage LLM pipeline. End-to-end functional but explicitly pre-launch, safety-critical content gated behind self-vetted review only.
RaiVaibhav /
agentgate
TypeScript MCP permission gateway with typed, documented architecture. Permission engine, response scanning, and audit logging. No tests/CI; experimental-stage product with solid technical foundation but nascent adoption (4 stars, 17 days old).
RaiVaibhav /
portfolio
Personal portfolio site built with Next.js 16, TypeScript, and static export. Well-designed, typed, and documented with two interactive demos (virtualized table, design tokens), but entirely single-repo, single-author, and launched same day (0 stars, 0 forks).
RaiVaibhav /
securespec
Claude skill scaffold for security-focused feature specs; 12KB single-push dump with README + design docs but no executable code, tests, or CI. One-day-old, experimental tool template.
RaiVaibhav /
RaiVaibhav
Personal portfolio/profile repo containing only a README with links and professional bio. No source code, no projects, no substance — purely a GitHub profile landing page.
RaiVaibhav /
warm-signal
Empty scaffold created Jul 31, 2026 with only README describing a hiring platform idea (vouch layer for job applications). Zero stars, no code, no git history, no license. Founders' diary entry, not a shipped project.
RaiVaibhav /
harbor-task
Empty scaffold with minimal git history (1 commit in 18 seconds), no README, no CI, untyped Python, and 3KB total size. No meaningful code or documentation present.
06 · Timeline
- Sep 18, 2016Joined GitHub
- Jun 21, 2026Created agentgate — Permission gateway for AI agents — enforce per-tool access control on any MCP server (GitHub, Stripe, Slack). Audit every decision. Revoke instantly.
- Jun 27, 2026Created securespec — A Claude skill that takes a vague feature idea and produces a code-ready, security-reviewed implementation spec. No more shipping broken access control because you forgot to think
- Jul 6, 2026Created harbor-task
- Jul 8, 2026Created bandhu — A light, companion-based mental health check-in app, Engineering docs- https://raivaibhav.github.io/bandhu/
- Jul 18, 2026Created NatureFix
- Jul 31, 2026Created warm-signal
- Aug 5, 2026Created RaiVaibhav
- Aug 27, 2026Created fredis — A learning project, meant to understand the internals of Redis by creating one
- Aug 30, 2026Created portfolio
- Aug 30, 2026Most recent push to portfolio
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.