▸ This tool was built by an AI agent from Zoral
← RATE MY GITHUB

#10 — Top 99.2%

andiradulescu

Andi Radulescu

A

Ship machine

Overall

0.0

/ 100

01 · Roasts

Linux Kernel Cosplayer

Your langPcts scream '92% C genius' but that's just Torvalds' 3.3GB repo sitting in your forks. Your actual original code is Python scripts and a CSS conference slide deck.

Documentation Maximalist, Test Minimalist

surroundpilot ships ARCHITECTURE.md, STATUS.md, design.md, CLAUDE.md, and README.md — but HAS_TESTS=no across every single original repo. You write more docs about tests than actual tests.

The 48-Hour Architect

abctl: 2 days. cc-plugin-codex: 2 days. surroundpilot: single-day push. visionpilot: 1-week burst. You design entire multi-daemon embedded systems in the time most engineers spend writing a ticket.

Follower Ratio Reality Check

111 followers, 701 following — that's a 6.3:1 following-to-follower ratio. You're engaging loudly with the community; it's just not fully reciprocating yet.

Stars: Critically Endangered

15 total stars across 109 repos. That's 0.14 stars per repo. The Linux kernel fork has more stars than everything you've actually written combined — and it has zero.

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

  • Impact
    25% weight
    95S
  • Consistency
    20% weight
    65C
  • Quality
    20% weight
    92S
  • Depth
    15% weight
    95S
  • Breadth
    10% weight
    40D
  • Community
    10% weight
    50D

03 · Stats

365-day commit heatmap

219 active days

Less
More

Language distribution

7 langs
  • C92%
  • C++4%
  • Python2%
  • Assembly1%
  • Shell0%
  • SWIG0%
  • Other1%

04 · Numbers

Owned repos

non-fork

23

Commits

last 12 months

194

Followers

111

Joined GitHub

Apr 2009

05 · Top repos

andiradulescu /

linux

95/100

This is the Linux kernel repository itself — an ecosystem-defining, multi-decade codebase representing the most critical open-source infrastructure project. 3.3GB codebase with sustained commits through 2026.

I95Q95D95
README
C01mo ago

andiradulescu /

cc-plugin-codex

40/100

Experimental plugin integrating Claude Code with Codex marketplace. Clear docs, tests, and typed Node.js runner script. Very new (2 days old), minimal stars, but shows structured design and deliberate feature scope.

I25Q60D35
READMETests
JavaScript22mo ago

andiradulescu /

visionpilot

40/100

Early-stage ROS 2 autonomous driving system integrating OpenPilot and Autoware paradigms for Rockchip NPU hardware (RK3588/RK3576). Comprehensive architecture docs and typed Python/Cython codebase, but minimal adoption (1 star), no tests/CI, and recent commits suggest active but pre-release development.

I25Q60D35
README
Unknown12mo ago

andiradulescu /

abctl

40/100

Niche low-level A/B boot control utility for Qualcomm UFS systems. Clear, well-documented single-file Python tool; typed-free but structured codebase with GPT manipulation and three UFS ioctl strategies. New (2 days old, 8 commits), no tests/CI.

I25Q60D35
README
Python02mo ago

andiradulescu /

surroundpilot

38/100

RK3588-based embedded ECU for surround-view localization with camera HAL, hardware acceleration (RGA/NPU), and UDP bridge to EnhancedOpenPilot. Well-structured with design docs, typed Python code, but extremely recent (5 days old) with minimal commits.

I25Q55D20
README
Unknown22mo ago

andiradulescu /

enhancedopenpilot

37/100

Specialized RK3588 fork of openpilot with hardware acceleration (NPU, RGA) and multi-camera HAL, featuring typed Python camera drivers and design docs, but early-stage with minimal adoption (4 stars, created 2026-03-17).

I25Q50D35
README
Unknown43mo ago

andiradulescu /

Context-is-All-You-Need

15/100

Minimal presentation-only repo for eu/acc Romania #3 event. 9 KB CSS with README and CI, but no substantive code, tests, or documentation beyond event metadata. One-off conference talk repo.

I15Q10D20
READMECI
CSS01mo ago

andiradulescu /

andiradulescu.github.io

7/100

Minimal GitHub Pages scaffold with only 1 KB repo size, 2 commits in 4 minutes, and a single-line README. No functional code, tests, CI, or documentation beyond greeting.

I5Q10D5
README
Unknown03mo ago

06 · Timeline

  1. Apr 23, 2009
    Joined GitHub
  2. May 10, 2024
    Created linux
  3. Feb 27, 2026
    Created andiradulescu.github.io
  4. Mar 10, 2026
    Created abctl
  5. Mar 10, 2026
    Created Context-is-All-You-Need — Presentation for eu/acc Romania #3: Agentic Engineering
  6. Mar 17, 2026
    Created enhancedopenpilot — Enhanced openpilot fork with RK3588 NPU acceleration and multi-camera tracking
  7. Mar 17, 2026
    Created surroundpilot
  8. Mar 17, 2026
    Created visionpilot
  9. Apr 2, 2026
    Created cc-plugin-codex — Use Claude Code from Codex to review code or delegate tasks. Opposite of openai/codex-plugin-cc.
  10. Apr 8, 2026
    Most recent push to Context-is-All-You-Need

07 · Compare

github.com/
andiradulescu · 6dmedian coder

08 · Rubric

How this score was produced

Overall = Σ (category × weight) + gentle top-end curve

CategoryWeightScoreContrib.
Raw total78.4
Top-end curve+5.2
Final overall83.6

Tier thresholds

S90100Mass-producing humansA8089Ship machineB7079Solid engineerC6069Getting thereD4059README enthusiastF039GitHub tourist
▸ How the pipeline works
  1. 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.
  2. 02Triage.A small model reads every repo's file tree + README and picks the 20 files per repo that actually reveal how you code.
  3. 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.
  4. 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.
  5. 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.
andiradulescu · 83.6/100 — Rate My GitHub