01 · Roasts
Sprint Merchant
11 repos with zero tests, zero CI, and at least 4 of them born and abandoned within a single afternoon. ev-cv was created AND last pushed within a 3-minute window. Three. Minutes.
The Ghost Heatmap
163 commits across 52 weeks, but the heatmap is ~90% zeros. Your entire annual output fits inside a few frantic weekends — the GitHub lawn looks like a parking lot after a hailstorm.
Economics Mono-Culture
econ-diagrams, ev-cv, ev-cv-3d, diamond-mirrlees-diagram, sg-carry-trade-blog — five repos all orbiting microeconomics. Bold domain commitment, or just one homework assignment that escaped containment?
README? Never Heard of Her
diamond-mirrlees-diagram, sg-carry-trade-blog, and payment-service all ship zero README. You wrote a METHODOLOGY.md for a debate simulator but couldn't type a single sentence for a payment service with known security bugs.
Lone Wolf With No Pack
soloPct = 100%, totalPRsYear = 0, totalIssuesYear = 0, followers = 3. GitHub is a social platform and you're using it as a personal USB drive.
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% weight65C
- Community10% weight25F
03 · Stats
365-day commit heatmap
25 active days
Language distribution
- Jupyter Notebook37%
- Python24%
- JavaScript16%
- TypeScript10%
- HTML9%
- CSS5%
04 · Numbers
Owned repos
non-fork
15
Commits
last 12 months
163
Followers
3
Joined GitHub
Oct 2024
05 · Top repos
t1mchee /
t1mchee.github.io
Personal website with hand-written HTML/CSS, interactive canvas background, and economics blog posts with interactive visualizations; well-documented project architecture but limited external visibility.
t1mchee /
this-house-believes-in-vibes
Multi-agent Cambridge Union debate simulator using LangGraph, persona-grounded RAG, and three-layer judging. Early-stage personal project (2 days old, 9 commits) with typed Python, extensive documentation (METHODOLOGY.md, docs/, ARCHITECTURE.md), and structured multi-file layout but no tests/CI and unclear practical im
t1mchee /
econ-diagrams
Niche single-author library (0 stars, ~900 KB) for declarative linear economics diagrams. Typed Python, structured, documented with README + AGENTS.md; ships with 2 working demos. Created and last pushed same day (2026-04-21, 3 commits), experimental artifact targeting LLM agents.
t1mchee /
ev-cv-3d
TypeScript React + Plotly visualization tool for 3D utility surfaces and indifference curves in microeconomics education. Companion to ev-cv project; ships with 7 interactive modes but no tests, CI, or license.
t1mchee /
ev-cv
Educational React+D3 visualization of microeconomic concepts (deadweight loss, tax incidence, welfare effects). Typed, documented, structured codebase with sophisticated economic calculations and interactive controls. Created and pushed same day (2026-04-22) with minimal commit depth.
t1mchee /
linear-regression-deep-dive
Educational Jupyter notebook implementing linear regression, Ridge, and Lasso from NumPy with visualizations. No stars/adoption, but well-documented README and structured content across 3 months of commits (3 of 30 recent).
t1mchee /
diamond-mirrlees-diagram
Educational interactive visualization of Diamond-Mirrlees production efficiency using D3.js with MathJax-rendered economics theory. Solo project, minimal documentation, no tests or CI.
t1mchee /
payment-service
Early-stage TypeScript payment service with intentional bugs for educational/demonstration purposes. Has typed code, tests, and clear error handling, but no documentation, no CI, no license, and serious unresolved bugs (null-check race condition, timing attack).
t1mchee /
sg-carry-trade-blog
Early-stage blog post + analysis repo using Python with matplotlib for financial visualizations. Unfinished: analysis.py is incomplete (truncated mid-function), no README, no tests, no CI, no documentation structure. 2 commits over 13 minutes on 2026-04-14.
t1mchee /
devinguard
Early-stage orchestration service connecting monitoring tools to Devin AI for autonomous incident response; untyped, no tests/CI, minimal code samples visible, 4-day-old project with sparse commit history.
t1mchee /
timchee.github.io
Personal GitHub Pages site with minimal documentation. 9.9 KB CSS codebase, 5 commits in last 30 days, no tests/CI/license. README is a bare title with no substantive content describing the project.
06 · Timeline
- Oct 15, 2024Joined GitHub
- Oct 29, 2024Created timchee.github.io
- Oct 30, 2025Created linear-regression-deep-dive — Comprehensive Jupyter notebook teaching linear regression (OLS, Ridge, Lasso) from scratch with NumPy. Includes implementations, visualizations, and interactive learning tools.
- Feb 19, 2026Created this-house-believes-in-vibes — Multi-agent Cambridge Union debate simulator — persona-grounded RAG + LangGraph + three-layer judging + Monte Carlo ensemble analysis
- Feb 25, 2026Created t1mchee.github.io — Personal website
- Mar 7, 2026Created devinguard
- Mar 7, 2026Created payment-service
- Apr 7, 2026Created diamond-mirrlees-diagram — Interactive diagram exploring the Diamond-Mirrlees production efficiency result
- Apr 14, 2026Created sg-carry-trade-blog — Blog post: Why the Singapore carry trade doesn't exist. Analysis of MAS's crawling band framework and its effect on carry trade profitability.
- Apr 21, 2026Created econ-diagrams — Declarative Python library for linear economics diagrams, built on matplotlib
- Apr 22, 2026Created ev-cv — Interactive visualization of equivalent variation, compensating variation, and deadweight loss
- Apr 22, 2026Created ev-cv-3d — 3D utility-surface visualisation (companion to ev-cv)
- Apr 23, 2026Most recent push to diamond-mirrlees-diagram
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.