01 · Roasts
The Humor Cinematic Universe
assignmen_3_humor_henry, assignmenthumor2, Assignment_2_humor, Assignmen_3_humor — four repos, one vibe, and a cumulative 0 stars. You've built the MCU of internal admin dashboards that nobody asked for.
UrAgent: Peak Software Engineering
UrAgent was created and pushed in under 2 seconds. That's not a repo, that's a sneeze. The README is empty, the code is empty, the ambition was apparently also empty.
94% Python, 0% README
Python owns 94% of your codebase yet half your repos have no README whatsoever. You clearly know how to write code. You just refuse to explain it to anyone including future you.
Commit Archaeology Required
Your heatmap is a desert for 40 straight weeks followed by a sudden oasis. 193 commits in a year but most of them happened in the last 3 months. Were you hibernating or just waiting for deadlines?
CI/CD: Consistently Ignored / Completely Disabled
Not a single repo across your entire profile has CI enabled. You have Vitest, Playwright, and pytest all configured — you know what tests are — but apparently the concept of running them automatically is too avant-garde.
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% weight55D
- Quality20% weight62C
- Depth15% weight50D
- Breadth10% weight55D
- Community10% weight25F
03 · Stats
365-day commit heatmap
58 active days
Language distribution
- Python94%
- TypeScript3%
- Cython1%
- CSS1%
- JavaScript0%
- Jupyter Notebook0%
- Other1%
04 · Numbers
Owned repos
non-fork
34
Commits
last 12 months
193
Followers
2
Joined GitHub
Aug 2024
05 · Top repos
HenryWashuHe /
assignmen_3_humor_henry
Next.js admin dashboard for managing LLM prompt pipelines ("humor flavors"). Typed, documented, multi-file architecture with Supabase, Tailwind, and drag-drop UI. No tests/CI, single author commit history, internal tool without external adoption signals.
HenryWashuHe /
assignmenthumor2
Personal humor-captioning web app built with Next.js, TypeScript, Supabase, and MediaPipe gesture detection. Typed, documented, and structured, but private (0 stars), minimal external adoption, and experimental scope.
HenryWashuHe /
2026DevFest
Educational auscultation AI built for WashU 2026 hackathon. Combines Raspberry Pi audio capture, PyTorch CNN classifier, and LangGraph multi-agent Gemini pipeline. Well-documented with typed Python backend, React frontend, and 9 commits in 6 days.
HenryWashuHe /
research_demo
Research project extracting and analyzing mathematical concepts from textbooks using SAE-based sparse feature analysis. Typed Python with clear modular structure and comprehensive docs, but nascent (14-day-old), zero external adoption, and no tests.
HenryWashuHe /
Assignment_2_humor
A typed Next.js admin panel for a humor study research project, with unit & e2e tests, Supabase integration, and structured components. Single-week sprint (10 days, 3 recent commits), no CI/CD pipeline, experimental scope.
HenryWashuHe /
NLP-HW3
NLP homework assignment implementing LLM-based tool-calling agents for retail customer service tasks. Contains student-completed agent implementations (LangChain, OpenAI SDK), domain tools with Pydantic validation, and integration tests hitting live APIs.
HenryWashuHe /
NLP-HW2
Homework submission implementing GPT-2 from scratch in PyTorch with training on 20 Newsgroups. Minimal stars/adoption, typed code and structured src/, but lacks tests, CI, documentation, and license.
HenryWashuHe /
COMS-4705-NLP-HW1
Coursework assignment repo for NLP/Word2Vec with 0 stars, no README, no tests/CI. 9 commits across 8 days (Feb 11-19, 2026). Implementation fills in student-marked sections of assignment starter code.
HenryWashuHe /
ML-Theory-PSET1
Single-file homework assignment implementing bandit algorithms (UCB, epsilon-greedy, arm elimination) with no documentation, tests, or CI. Appears to be a one-off tutorial exercise with minimal git history (2 commits in last 30 days).
HenryWashuHe /
Assignmen_3_humor
Minimal assignment submission with no stars, zero documentation, no tests/CI, and only 2 commits in 25 minutes on 2026-03-23. Pure scaffold/tutorial exercise.
HenryWashuHe /
aws_bedrock_withRAG
Minimal experimental RAG chatbot prototype with AWS Bedrock and Llama3. One-off dump with 2 Python scripts, no README, tests, CI, or documentation. Created and last pushed same day (2026-03-12).
HenryWashuHe /
UrAgent
Empty scaffold repository with zero files, no documentation, and no commits beyond initialization. Created and pushed within seconds on 2026-03-06 with no meaningful content.
06 · Timeline
- Aug 24, 2024Joined GitHub
- Jan 31, 2026Created assignmenthumor2
- Feb 11, 2026Created COMS-4705-NLP-HW1
- Feb 22, 2026Created research_demo
- Mar 4, 2026Created Assignment_2_humor
- Mar 6, 2026Created UrAgent
- Mar 12, 2026Created aws_bedrock_withRAG
- Mar 20, 2026Created NLP-HW2
- Mar 23, 2026Created Assignmen_3_humor
- Mar 23, 2026Created assignmen_3_humor_henry
- Apr 2, 2026Created NLP-HW3
- Apr 4, 2026Created ML-Theory-PSET1
- Apr 6, 2026Created 2026DevFest
- Apr 19, 2026Most recent push to NLP-HW3
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.