01 · Roasts
The 1-Hour MVP Factory
louder-assignment: 7 commits in 60 minutes. vision-agent-hack: 13 commits in 3 hours. HRMS-Lite: 6 commits in 10 hours. Your development philosophy appears to be 'ship it before the adrenaline wears off, never return.'
90% Python, 0% Tests
Python is 90% of your codebase and exactly 1 of your 10 scored repos has a test suite—and that one is written in Julia. Your primary language is your most untested language.
The Heatmap Black Hole
Your activity heatmap has ~22 consecutive weeks of near-zero commits (weeks 11–32). That's roughly half a year where GitHub forgot you existed. The recent sprint in weeks 39–44 is appreciated, but the hibernation is hard to unsee.
meta-rl-status200: A Study in Ambition
You created a repo called meta-rl-status200, pushed 1 commit, wrote a README that says only 'meta-rl-status200', and walked away. At least the HTTP status code is aspirational.
106 Repos, 42 Total Stars
With 106 public repos you're averaging 0.4 stars each. The profile README has more stars (2) than most of your actual projects. Quantity is not the strategy it appears to be.
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% weight72B
- Depth15% weight50D
- Breadth10% weight45D
- Community10% weight50D
03 · Stats
365-day commit heatmap
67 active days
Language distribution
- Python90%
- C++4%
- Cython2%
- HTML2%
- Jupyter Notebook1%
- C1%
04 · Numbers
Owned repos
non-fork
41
Commits
last 12 months
183
Followers
39
Joined GitHub
Jun 2021
05 · Top repos
calebjubal /
AhoCorasick.jl
Polished Julia implementation of Aho-Corasick algorithm with comprehensive tests, CI, and clear documentation. Early-stage personal project with no adoption signals yet.
calebjubal /
louder-assignment
Fresh full-stack SaaS app (Next.js + FastAPI) for AI-driven event planning. Typed, documented, deployed, but nascent—created 2026-03-20 with only 7 commits in 1 hour. No tests, no CI, no stars/forks. Demonstrates competent architecture but pre-portfolio maturity.
calebjubal /
vision-agent-hack
Personal AI agent project using Vision Agents framework to provide a Joey Tribbiani-themed frontend development assistant. 741 KB codebase with structured Python setup, comprehensive README, and multimodal integration (Gemini, ElevenLabs, Deepgram), but no tests, CI, or type hints despite Python being the primary langu
calebjubal /
resume-latex-updater
TypeScript Next.js resume editor with LaTeX generation, Clerk auth, and Neon database. Clean architecture, structured components, and type-safe API routes, but nascent project (created 2026-03-26, 10 commits in 2 hours).
calebjubal /
HRMS-Lite
A lightweight HRMS project built with Flask backend + React 19 frontend. Has README, typed Pydantic schemas, clean API structure, but lacks tests, CI, license, and is only 56 KB with 6 commits in a single day—experimental stage.
calebjubal /
multi-doc-rag
Early-stage RAG chatbot built with Streamlit, LangChain, and Cerebras API. Minimal codebase (12 KB) with working PDF processing and conversational retrieval, but lacks tests, CI, and substantive documentation. Created 14 days ago with 8 commits showing initial development burst.
calebjubal /
IITD_Feb26_AAIPL
Experimental MCQ generation pipeline using local LLMs (Phi4) to create/answer questions. Unfinished, minimal README, no tests/CI/license. Created Feb 2026, 22 KB with ~8 commits in 1 hour. Lacks documentation and production-ready structure.
calebjubal /
calebjubal
GitHub profile config repo with only a decorative README. 8 KB, 2 stars, no code, no tests, no CI, no license. Pure cosmetic portfolio piece with no executable substance.
calebjubal /
django-test
Bare scaffold repo with minimal content (4 KB), created 2026-01-29 with only 3 commits in 2 minutes. README is title-only, no source code sampled, no tests, no meaningful documentation despite HAS_README=yes flag.
calebjubal /
meta-rl-status200
Empty scaffold repo with only a bare README title. Created moments ago (2026-04-02) with 1 commit. No code, tests, CI, license, or meaningful documentation. Placeholder status.
06 · Timeline
- Jun 3, 2021Joined GitHub
- Jun 5, 2021Created calebjubal — Config files for my GitHub profile.
- Jan 22, 2026Created multi-doc-rag
- Jan 29, 2026Created django-test
- Feb 6, 2026Created HRMS-Lite
- Feb 9, 2026Created AhoCorasick.jl
- Feb 15, 2026Created IITD_Feb26_AAIPL
- Mar 1, 2026Created vision-agent-hack
- Mar 20, 2026Created louder-assignment
- Mar 26, 2026Created resume-latex-updater
- Apr 2, 2026Created meta-rl-status200
- Apr 21, 2026Most recent push to AhoCorasick.jl
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.