01 · Roasts
One-star constellation
33 public repos, 1 total star, and 1 follower: the portfolio is shipping faster than its audience is arriving.
Sprint specialist
Haven and open-transcribe show real scope, but both demonstrate two-day build trajectories rather than long maintenance arcs.
CI left on read
Haven, dotclaude, and open-transcribe all skip CI; the automation layer never made the commit.
Night owl, scattered calendar
100% night-owl activity is committed, but the heatmap still contains long blank stretches around those bursts.
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% weight30F
- Consistency20% weight55D
- Quality20% weight62C
- Depth15% weight50D
- Breadth10% weight55D
- Community10% weight25F
03 · Stats
365-day commit heatmap
83 active days
Language distribution
- TypeScript76%
- JavaScript14%
- Python3%
- CSS2%
- Ruby2%
- HTML1%
- Other2%
04 · Numbers
Owned repos
non-fork
14
Commits
last 12 months
302
Followers
1
Joined GitHub
Sep 2016
05 · Top repos
contactkvijay /
haven
A substantial, typed Next.js/SQLite household-document app with unusually careful insurance-domain modeling, privacy boundaries, extraction and audit logic, but only 1 star and a two-day trajectory limit demonstrated adoption.
contactkvijay /
open-transcribe
A substantial weekend-built Chrome extension plus FastAPI service for video transcription and X bookmark archiving, with clear setup documentation and focused tests, but no demonstrated adoption yet.
contactkvijay /
dotclaude
A focused Windows-only Claude Code slash command that turns clipboard screenshots into timestamped PNGs, with a polished README and MIT license but no tests, CI, or demonstrated adoption.
06 · Timeline
- Sep 19, 2016Joined GitHub
- May 1, 2026Created open-transcribe
- May 13, 2026Created dotclaude — Claude Code slash commands, skills, and hooks. Reusable prompts for Claude Code CLI and Codex CLI. MIT licensed.
- Aug 9, 2026Created haven — A safe place for your family's papers
- Aug 11, 2026Most recent push to haven
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.