01 · Roasts
The 9-Minute Architect
claude-code-skills was 'created' in 9 minutes, claude-onboard in 3 minutes, token-surgen in 5. Bro is shipping repos faster than GitHub can render the README.
CI/CD Who?
8 repos scored, 0 have CI. You added a CONTRIBUTING.md checklist to claude-code-skills that asks for quality bars you don't enforce anywhere in your own code.
papermarket Is Offline
Your most technically impressive project — multi-service React+Node+Postgres with Polymarket sync — is 'currently offline'. The vibes are there, the uptime is not.
Burst Season, Then Nothing
Your heatmap is 30+ weeks of near-zero commits followed by weeks 34–45 of wall-to-wall 4s. You're not a developer, you're a seasonal migration pattern.
Solo 100%, Community 0%
soloPct is literally 100%. In 14 PRs and 4 issues this year, there is zero evidence you've touched anyone else's code. GitHub is multiplayer, Nathan.
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% weight56D
- Consistency20% weight60C
- Quality20% weight57D
- Depth15% weight55D
- Breadth10% weight65C
- Community10% weight50D
03 · Stats
365-day commit heatmap
116 active days
Language distribution
- Python59%
- TypeScript33%
- JavaScript3%
- CSS2%
- HTML2%
- Lua0%
- Other1%
04 · Numbers
Owned repos
non-fork
31
Commits
last 12 months
274
Followers
10
Joined GitHub
Feb 2025
05 · Top repos
natedemoss /
papermarket
Full-stack paper trading prediction market platform (React + Node.js + Postgres) with Polymarket data sync. Early-stage active project with typed code, tests, documentation, and multi-service architecture, but currently offline and limited adoption (16 stars).
natedemoss /
Teammind
TypeScript Claude Code integration tool with git-aware team memory capture, local semantic search via HuggingFace embeddings, and MCP server. 225KB codebase, 25/30 recent commits, ships with typed code, README, tests absent, CI absent.
natedemoss /
Claude-Code-Wrapped-Skill
Claude Code skill that generates an interactive terminal slideshow of your AI coding statistics inspired by Spotify Wrapped. Typed Python 3.8+, well-documented README with clear installation and usage instructions, structured single-file implementation with caching, multi-slide visualization system. No tests or CI. ~87
natedemoss /
claude-code-skills
Early-stage Claude Code skills collection with 3 working interactive skills (wrapped, token-surgeon, onboarding). Well-documented with clear structure and MIT license, but brand new (9-minute creation window) with minimal adoption and no tests/CI.
natedemoss /
Chorus
Early-stage Electron + TypeScript desktop tool for parallelizing Claude Code agents via git worktrees. Typed, documented, structured with React/Vite frontend, but brand new (1 star, 6 commits in 15 minutes) with no tests or CI.
natedemoss /
natedemoss
GitHub profile config repo with personal bio and project portfolio. Zero stars, minimal technical depth, but README showcases 5+ named projects suggesting active shipping pattern.
natedemoss /
claude-onboard
Claude Code skill for AI-assisted onboarding guide generation; brand-new (2h old), 4KB, no source files, 2 commits. Abandoned prototype or placeholder setup.
natedemoss /
token-surgen-claude-skill
Claude Code Skill for analyzing token waste in prompts with 10 named patterns. Single-day creation (2026-03-23), 4 commits, 10 KB scope, MIT licensed. README documents the concept but no source files available for review.
06 · Timeline
- Feb 6, 2025Joined GitHub
- Feb 6, 2025Created natedemoss — Config files for my GitHub profile.
- Mar 21, 2026Created papermarket — Paper-Trading prediction market
- Mar 23, 2026Created Claude-Code-Wrapped-Skill — A Claude Code skill that returns a slideshow in the CLI of your Claude Code "Wrapped" inspired by Spotify Wrapped.
- Mar 23, 2026Created token-surgen-claude-skill — A Claude Code Skill that analyzes prompts and gives feedback on how to save tokens.
- Mar 23, 2026Created claude-onboard
- Mar 23, 2026Created claude-code-skills — A collection of Claude Code skills
- Mar 25, 2026Created Teammind — Git-aware team memory for Claude Code
- Apr 22, 2026Created Chorus — Run many Claude Code agents in parallel in isolated git worktrees. Pick the winner.
- Apr 29, 2026Most recent push to Claude-Code-Wrapped-Skill
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.