01 · Roasts
TypeScript or Bust
97% TypeScript across 51 repos. Impressive commitment to a single language — or a deep fear of touching anything else. Python got 0%, HTML got 0%, and CSS is technically in your portfolio the same way a participation trophy is an award.
CI? Never Heard of Her
Out of 8 analyzed repos, only 2 have CI pipelines. You wrote tests for PickUp and NeoynGPT — then apparently forgot CI was an option for every other project. Your future self will remember this during a production incident.
The One-Day Architect
PickUp has X11 backends, Wayland backends, systemd daemons, VS Code integrations, tmux integrations, AND 12 test cases — all built in 4 commits spanning 49 minutes on June 8th. Either you move at superhuman speed or that's some very well-prompted boilerplate.
5 Followers, Building in the Void
cli-share has VT100 terminal emulation and a published npm package. NeoynGPT has an ARCHITECTURE.md. You have 5 followers. The work is real; the audience has not arrived yet — or is entirely in your private repos.
142 Public Commits, Plausibly More
142 public commits this year with privateWorkLikely=true — which means the server detected you're doing real work that GitHub can't see. You're building plany.space, shipping cli-share, and running hackathons... and only 6 PRs on other people's code all year.
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% weight56D
- Consistency20% weight60C
- Quality20% weight62C
- Depth15% weight55D
- Breadth10% weight45D
- Community10% weight25F
03 · Stats
365-day commit heatmap
105 active days
Language distribution
- TypeScript97%
- JavaScript2%
- Python0%
- HTML0%
- CSS0%
- Jupyter Notebook0%
- Other1%
04 · Numbers
Owned repos
non-fork
39
Commits
last 12 months
142
Followers
5
Joined GitHub
Apr 2023
05 · Top repos
saurav-shakya /
NeoynGPT-video-to-code
JavaScript agent skill for AI-driven video-to-code analysis via Gemini API. Typed-adjacent (dynamic JS), documented via SKILL.md + ARCHITECTURE.md, tests + CI present, MIT licensed. Novel MVP at 5 stars with 6 recent commits and ~1.3 MB codebase.
saurav-shakya /
cli-share
Working CLI tool with VT100 terminal emulation for sharing terminal sessions to GitHub Gists. Typed (JSDoc), documented README, structured monorepo, but no tests/CI and small initial release (6 stars, 1.0.5 published).
saurav-shakya /
PLANY-LANDING
Next.js landing page for Plany.space with TypeScript, component-driven architecture, test coverage (HAS_TESTS=yes), and substantive pricing/feature logic. Zero external adoption; solely promotional website for proprietary mobile app.
saurav-shakya /
PickUp
Brand-new Python utility for Linux session restore with dual X11/Wayland support, multi-app architecture, and systemd integration. Well-structured but very early (4 commits in 1 day, 0 external adoption signals).
saurav-shakya /
saurav-shakya.github.io
Personal portfolio website (56 KB) with README highlighting founder background and external projects (Plany.Space, cli-share). No source files sampled; minimal technical depth visible in this repo itself—serves as landing page linking to work elsewhere.
saurav-shakya /
DehliNcr_StartupMap
One-week TypeScript/React map explorer for Delhi NCR startups. Typed, structured Vite setup with Tailwind, but minimal tests/CI, thin docs, and fresh commit history (2 of last 30).
saurav-shakya /
SIH2026
Fresh Next.js boilerplate (36h old) for Smart India Hackathon 2026, TypeScript-based dashboard with mock telemetry simulation. No production use or external adoption signals yet.
saurav-shakya /
saurav-shakya
GitHub profile configuration repo with personal bio and project links. 14 KB, no source code, minimal utility beyond portfolio presentation.
06 · Timeline
- Apr 28, 2023Joined GitHub
- Sep 12, 2023Created saurav-shakya — Config files for my GitHub profile.
- Dec 22, 2024Created NeoynGPT-video-to-code — world's first Video to code harness skill
- Jan 2, 2026Created saurav-shakya.github.io — personal web
- Apr 19, 2026Created cli-share — share your claudecode ,codex or other cli tools terminal conversation via gist
- Jun 8, 2026Created PickUp — Session restore for Ubuntu — PickUp fixes that with one command.
- Jun 25, 2026Created PLANY-LANDING — Landing page of Plany.space
- Aug 15, 2026Created DehliNcr_StartupMap — google map for Startups
- Aug 16, 2026Created SIH2026
- Aug 26, 2026Most recent push to saurav-shakya
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.