01 · Roasts
The Burst Shipping Disorder
blink-2d, blink-3d, personal-job-autofill, lorentz-boost-shader, and github-blogger-agent were all last pushed on 2026-04-15. You don't have a coding habit — you have a coding episode.
senior-project Ends Mid-Sentence
Your high-school neural net sim literally cuts off at `ang = a` on line ~340. You submitted an incomplete sentence to GitHub and called it a senior project. The simulator can't run, the altruism model remains unresolved, and BASIS Independent Fremont is probably fine with that.
0 Stars, 0 Forks, 0 Followers
9 repos, totalStars=0, followers=0. The whole portfolio exists in a perfect social vacuum. You're shipping into the void with the confidence of someone who doesn't know GitHub has a social layer.
data_science_experiments: Hardcoded Desktop
`/Users/prsaketh/Desktop/...` is plastered into student_stats_APR18.py. This works on exactly one machine in the world: yours. Not a repo — a backup.
22 Public Commits in a Year
22 commits across 12 months is roughly one commit per fortnight. The heatmap is basically a calendar with two weeks circled in pencil. Private work might save you, but publicly this is spectator-tier participation.
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% weight55D
- Quality20% weight62C
- Depth15% weight50D
- Breadth10% weight65C
- Community10% weight25F
03 · Stats
365-day commit heatmap
8 active days
Language distribution
- Python66%
- Astro19%
- JavaScript7%
- HTML2%
- TypeScript2%
- CSS2%
- Other2%
04 · Numbers
Owned repos
non-fork
8
Commits
last 12 months
22
Followers
0
Joined GitHub
Aug 2022
05 · Top repos
sakethpilli007 /
sakethpilli007.github.io
Personal portfolio site built with Astro + Tailwind, shipping a polished comic-styled design system with structured content collections and CI/CD deployment. Early-stage but complete, small repo (2264 KB) with 1 commit in ~9 hours.
sakethpilli007 /
blink-3d
Experimental first-person stealth game with webcam blink control and Panda3D rendering. Typed Python, structured multi-module layout, documented README with controls; but zero stars/forks, ~6 commits in 1.5 months, no tests/CI/license.
sakethpilli007 /
blink-2d
One-week game jam project: Python Pygame stealth game with webcam blink detection, procedural level generation, enemy AI, and crafting. No tests or CI; untyped Python; limited architectural documentation; feature-complete but experimental in scope.
sakethpilli007 /
personal-job-autofill
Functional Chrome extension for auto-filling job applications with local profile storage; zero network requests, typed UI logic, and comprehensive field matching. Day-old repo with minimal commit history but complete feature set and structured multi-file architecture (popup.html, content.js, manifest.json, etc.).
sakethpilli007 /
lorentz-boost-shader
Single-file GLSL shader visualizing relativistic Lorentz transformations via ray tracing, with clear mathematics but minimal scope, no tests/CI, and only one commit in a single day.
sakethpilli007 /
github-blogger-agent
Flask app leveraging Gemini API to auto-generate GitHub repo reports from diffs and commit history. Typed in Python but unpolished: no tests, no CI, no license, thin docs, and only 4 of last 30 commits with 15 KB codebase suggest early-stage experiment.
sakethpilli007 /
senior-project
High-school senior project: incomplete neural-network ecosystem simulator for studying altruism emergence. Code is cut off mid-implementation, lacks tests/CI, incomplete. Tutorial-grade artifact.
sakethpilli007 /
data_science_experiments
One-off data science learning experiment with hardcoded paths, no documentation, tests, or CI. Created and last pushed on 2026-04-18 within hours apart (2 commits total).
06 · Timeline
- Aug 5, 2022Joined GitHub
- Oct 13, 2025Created github-blogger-agent — Flask app that generates AI-powered project reports and changelogs from any GitHub repo using Gemini.
- Feb 28, 2026Created blink-3d — First-person stealth/horror game — your blinks control reality. Built with Panda3D.
- Apr 15, 2026Created blink-2d — Top-down stealth maze game with webcam blink detection. Built with Pygame.
- Apr 15, 2026Created personal-job-autofill — Chrome extension that fills job application forms using locally stored profile data. 100% local, zero network requests.
- Apr 15, 2026Created lorentz-boost-shader — GLSL fragment shader visualizing relativistic length contraction via Lorentz-transformed ray tracing.
- Apr 18, 2026Created data_science_experiments
- Apr 21, 2026Created senior-project
- Apr 21, 2026Created sakethpilli007.github.io — Personal portfolio — PR Saketh, CS @ Cambridge.
- Apr 21, 2026Most recent push to sakethpilli007.github.io
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.