const aka = {
  name: "Aka",
  username: "akaanakbaik",
  location: "Sumatra Barat, Indonesia 🇮🇩",
  role: "Student Developer",
  focus: ["Web Development", "Linux", "Automation", "AI", "Bots", "Cloud"],
  motto: "still water, flowing. — air tenang menghanyutkan"
};

I build practical projects while learning: web apps, bot systems, Linux/VPS tools, automation, and AI-powered experiments.

       

This section is generated from the GitHub REST and GraphQL APIs. The verified snapshot runs automatically every hour at minute 17 UTC, on pushes that affect the profile, and on manual dispatch. Badge delivery uses a short cache to surface each completed snapshot promptly.

Language shares use the GitHub /languages byte counts for every public repository. The charts include concentration, entropy, Gini, coefficient of variation, and Pareto calculations instead of estimates from repository names.

   

Full Code Census. Every tracked, regular source file on the default branch of every public repository is read without a file-size limit, including tracked dependencies and generated source when present. “Lines” means non-empty code lines; “characters” means Unicode code points; notebooks contribute only their code cells. Empty files are counted as files with zero lines and characters. Symlinks do not contain source themselves and are recorded separately. Any unreadable, binary, or unparsable source file fails the snapshot rather than creating a partial total; the workflow then independently recounts every cached clone before publication.

Streaks come from the official contribution calendar. Commit hours use the complete owned public commit history returned by the API collector, converted to Asia/Jakarta, with known automation commits excluded so scheduled updates do not inflate personal activity. Any incomplete API collection aborts publication instead of displaying a partial metric.

 

The clock and quote are refreshed as part of the same hourly snapshot. The README therefore reports a verified publication snapshot rather than claiming second-by-second updates from a static file.

Ranking combines logarithmically scaled stars, forks, watchers, pull requests, issues, discussions, owned commits, and exponential push recency. The score is a transparent activity heuristic, not a popularity claim.

   

The rolling 30-day report is generated inside the same hourly snapshot as the other profile assets, with every activity date normalized to Asia/Jakarta so its badges and chart share one consistent publication point.

Working on   → Music Player Web App, Portfolio v2, Linux tools, AI tools, and bot systems
Learning     → React, Node.js, TypeScript, Docker, Cloud, DevOps, and AI Agents
Ask me about → JavaScript, Web Development, UI/UX, Linux, VPS, bots, and automation

_{Made with ❤️ by Aka}