01 · Roasts
Portfolio, not a fan club
128 public repos and 103 total stars: the shipping cadence is real, but adoption has not caught up.
CI enjoys the spotlight
orphanode, coreloop, unixdm, and evalcore-infra are rigorously checked; resume-pdf and the profile README brought vibes instead.
Notebook monopoly
Jupyter Notebook accounts for 92% of language bytes, even while Rust, Go, C++, and Kotlin projects try to diversify the set.
Horizontal builder unlocked
215 multi-repo recent commits is serious output volume; now give the best tools enough runway to earn users.
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% weight68C
- Consistency20% weight65C
- Quality20% weight81A
- Depth15% weight58D
- Breadth10% weight65C
- Community10% weight55D
03 · Stats
365-day commit heatmap
298 active days
Language distribution
- Jupyter Notebook92%
- C++3%
- C2%
- Rust2%
- Go1%
- Shell1%
04 · Numbers
Owned repos
non-fork
46
Commits
last 12 months
327
Followers
81
Joined GitHub
Apr 2018
05 · Top repos
wthrajat /
unixdm
A polished early-stage C++17 terminal download manager with parallel range transfers, resumable atomic checkpoints, retry handling, TUI/plain modes, integration fixtures, and strong cross-platform CI, but only 2 stars and no demonstrated external adoption.
wthrajat /
coreloop
Coreloop is a well-engineered, documented Go/TypeScript Telegram learning service with durable jobs, multi-source ingestion, AI enrichment, security controls, tests, and CI, but it has only 2 stars and no demonstrated external adoption.
wthrajat /
evalcore-infra
A documented TypeScript/Python evaluation platform with FastAPI, Next.js, Celery, PostgreSQL, Redis, ingestion, scoring, adjudication, analytics, exports, tests, and dual CI workflows; adoption is not yet evidenced by its 0 stars and 0 forks.
wthrajat /
orphanode
Orphanode is a young Rust CLI with unusually rigorous static reachability analysis, deterministic reports, conservative fixes, workspace support, and broad automated validation, but it has only 2 stars and no demonstrated external adoption.
wthrajat /
lightning-maps
Lightning Maps is a documented, typed Next.js educational simulator with deterministic 50-node routing, failure scenarios, payment animation, metrics, experiments, and substantial automated tests, but has no visible adoption or CI/license.
wthrajat /
autocommit-rs
A polished Rust CLI that generates, validates, caches, and interactively applies Conventional Commits through OpenAI or Gemini, with CI-driven formatting and Clippy checks but limited visible adoption.
wthrajat /
bisn-cooling-curve-analysis
A substantial, reproducible Bi–Sn experimental analysis package with auditable raw-data handling, generated reports, figures, and workbooks, but currently has no visible adoption, tests, CI, or typed-language signal.
wthrajat /
android-auto-torque
AA Torque is a substantial, documented Kotlin Android Auto dashboard integrating Torque Pro, with gauges, charts, themes, persistence, crash reporting, and automated APK releases, but currently has 1 star and no tests or demonstrated external adoption.
wthrajat /
homebrew-unixdm
A focused Homebrew tap packaging UnixDM with a documented install path, pinned formula source, formula-level smoke tests, and two carefully pinned automation workflows, but no visible adoption or standalone test suite.
wthrajat /
tag-experiment
A small typed TypeScript GitHub-tag experiment with a documented setup, semver filtering, and a workflow, but no adoption and largely placeholder testing/build automation.
wthrajat /
resume-pdf
A narrowly scoped generated résumé artifact for cv.wthrajat.com, with a minimal README and no sampled source files, tests, CI, license, or typed implementation evidence.
wthrajat /
wthrajat
A low-adoption GitHub profile README centered on generated stats and animated media, with 4 stars and no demonstrated software package, tests, CI, or typed source tree.
06 · Timeline
- Apr 24, 2018Joined GitHub
- Oct 25, 2022Created wthrajat — Readme for chhapri teens
- Jun 19, 2025Created tag-experiment — Fetch release tags metadata from Github repo URL
- May 11, 2026Created unixdm — Internet Download Manager 🥀 (IDM) but for unix and unix-like systems ;)
- May 11, 2026Created homebrew-unixdm
- May 22, 2026Created autocommit-rs — Auto commit with conventional messages for free :)
- Jun 7, 2026Created android-auto-torque — Performance Monitor for cars with Android Auto
- Jul 3, 2026Created evalcore-infra — Create, orchestrate, review and export LLM evaluation datasets
- Jul 27, 2026Created resume-pdf — Generated public résumé artifact for cv.wthrajat.com
- Aug 3, 2026Created coreloop — private, self-hosted, telegram-first technical curriculum for efficient and easy learning
- Aug 16, 2026Created bisn-cooling-curve-analysis — cooling curves for bismuth–tin alloys
- Aug 17, 2026Created orphanode — find unused everything in your nodejs application
- Aug 26, 2026Created lightning-maps
- Sep 2, 2026Most recent push to coreloop
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.