01 · Roasts
Star-to-Repo Ratio Speedrun
346 public repos, 7998 total stars — that's ~23 stars per repo on average. Your GitHub profile is less a portfolio and more a very well-organized attic. gunshi has 403; the rest are fighting over the scraps.
WIP Eternal
ox-jsdoc is 3 weeks old and already has 4 crates, NAPI bindings, WASM output, and a cross-language parity test. Either you move impossibly fast or you start ambitious projects before the previous ones finish. Narrator: it was the latter.
Config Package Hoarder
eslint-config, eslint-plugin, prettier-config, tool-chain-config-checker — you've written more tooling to enforce your own opinions than most teams use in production. At some point the scaffolding IS the project.
34% Graveyard Ratio
staleRepoRatio=0.34 means roughly 1-in-3 of your repos haven't been touched in 2+ years. With 346 repos, that's ~117 digital fossils. The @kazupon GitHub account is 40% museum, 60% active lab.
Vue Core Member, Benchmarks Everything
args-tokens is 2 months old and already has mitata + vitest bench, a design.md, an ARCHITECTURE.md, and a STATUS.md. The README is longer than most developers' entire project histories. Touch grass. Ship v1.
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% weight75B
- Quality20% weight77B
- Depth15% weight72B
- Breadth10% weight72B
- Community10% weight55D
03 · Stats
365-day commit heatmap
328 active days
Language distribution
- TypeScript64%
- JavaScript18%
- Rust14%
- HTML1%
- Vue1%
- Shell0%
- Other2%
04 · Numbers
Owned repos
non-fork
65
Commits
last 12 months
1,576
Followers
4,300
Joined GitHub
Apr 2009
05 · Top repos
kazupon /
gunshi
Modern TypeScript CLI library with 403 stars, comprehensive type safety, extensive plugin system, full test coverage via vitest, CI/CD, and production-quality documentation. Monorepo with 6+ sub-packages targeting Node.js/Deno/Bun runtimes.
kazupon /
args-tokens
High-performance CLI argument parser with Node.js parseArgs token compatibility. Typed TypeScript, comprehensive documentation (README + design docs), full test/CI suite, benchmarked 2.5x faster than Node's util.parseArgs; early-stage project (1338 KB, 25 stars, created Feb 2025).
kazupon /
jts-utils
Well-crafted TypeScript utility library with comprehensive types, tests, and CI. Covers Result/Option, event emitters, observers, and string/type utilities. Low adoption (5 stars) but solid craftsmanship suggests active portfolio project by experienced developer.
kazupon /
eslint-plugin
Niche TypeScript ESLint plugin with 4 focused comment linting rules. Fully typed, well-documented (README + design docs), comprehensive test suite (vitest), and strict CI pipeline. 930 KB codebase shows meaningful depth, though minimal adoption (6 stars).
kazupon /
tool-chain-config-checker
TypeScript monorepo providing opinionated checkers for tsconfig and Vite configs with full test coverage and CI/CD. Narrow scope (2 packages, 231 KB), minimal adoption (2 stars), but well-structured and documented.
kazupon /
textlint-rule-no-pair
Specialized textlint rule for detecting disallowed bracket pairs, with comprehensive built-in pair support, full TypeScript typing, CI/CD, and multi-language test coverage but minimal adoption (0 stars).
kazupon /
eslint-config
Personal ESLint config package with 20+ modular presets (TypeScript, Vue, React, etc.), typed, tested, CI-enabled. Small adoption (9 stars) but complete, shipped tooling.
kazupon /
jsr-exports-lint
TypeScript linter for JSR exports field validation in tsdown/unbuild. Well-crafted, fully tested, and documented but narrow niche use case with minimal adoption (5 stars, no external integrations beyond peer dependencies).
kazupon /
ox-jsdoc
WIP Rust JSDoc parser inspired by oxc, with multi-language bindings and cross-language parity tests. Typed, structured, and well-documented but explicitly unstable for production use.
kazupon /
textlint-rule-no-period-on-paragraph-end
TypeScript textlint rule with full documentation, tests, CI, and MIT license. Specialized linting rule for paragraph-end punctuation. Only 1 star but complete tooling for indie project.
kazupon /
vue-i18n-continuous-localization-demo
Demo project for vue-i18n continuous localization with CI/CD setup, minimal stars (3), but consistent recent activity and structured tooling (tests, CI, license).
kazupon /
prettier-config
Personal Prettier config package with minimal adoption (1 star), no source files visible. Has CI and license but lacks tests and meaningful architectural depth.
06 · Timeline
- Apr 12, 2009Joined GitHub
- Dec 21, 2019Created vue-i18n-continuous-localization-demo
- Feb 21, 2024Created tool-chain-config-checker — An optionated JS / TS tool chain config cheker
- May 22, 2024Created textlint-rule-no-period-on-paragraph-end — textlint rule for no period on paragraph end
- May 27, 2024Created textlint-rule-no-pair — textlint rule for no pair
- Jun 30, 2024Created prettier-config — Prettier config for @kauzpon
- Jul 2, 2024Created eslint-config — ESLint config for @kazupon
- Aug 24, 2024Created jts-utils — Opinionated collection of JavaScript / TypeScript utilities by @kazupon
- Feb 11, 2025Created args-tokens — parseArgs tokens compatibility and more high-performance parser
- Feb 24, 2025Created gunshi — Gunshi: Modern javascript command-line library
- Apr 6, 2025Created jsr-exports-lint — JSR exports field lint with tsdown
- Apr 7, 2025Created eslint-plugin — ESLint plugin for @kazupon
- Apr 4, 2026Created ox-jsdoc — High performance jsdoc parser
- Apr 27, 2026Most recent push to textlint-rule-no-pair
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.