I build local tools for reviewing performance, CI configuration, and coding-agent instructions. Current focus: making saved benchmark runs easier to compare.
Two Hyperfine JSON exports → a before/after comparison, configurable regression budget, and Markdown you can copy into a review.
Get the browser ZIP — no installation · English quick start · راهنمای فارسی
The comparison runs locally. It highlights mean changes; it does not establish statistical significance.
- Agent Rule Conflicts — inspect coding-agent instructions for potential contradictions.
- MatrixCart — expand static GitHub Actions matrices, including
include/exclude, before spending runner time.
Try a tool on your own workflow and tell me where it falls short through its issue tracker. Follow for releases and concrete improvements to these tools.
فارسی: تمرکز فعلی روی بهترشدن همین ابزارهای قابلاستفاده است. تجربهٔ استفاده و مثال کوچکِ بدون اطلاعات حساس را در Issues پروژه بنویسید.
Note
My public tools favor deterministic behavior, local-first workflows, explicit limitations, and outputs that are easy to review or automate.
| Focus | Systems programming and developer tooling |
| Languages | Rust · C · C++ · Go · Python · R · Assembly |
| Approach | Small scope, explicit limits, tests, and CI |
| Status | ✅ Building in public |
🧭 Engineering loop
flowchart LR
A["Real workflow friction"] --> B["Small dependable tool"]
B --> C["Tests and CI"]
C --> D["Reviewable open source"]
I work across systems programming, developer tooling, automation, and scientific computing. I enjoy understanding how software behaves below the abstraction layer, then turning that knowledge into reliable, maintainable tools.
- Contributing focused fixes and tests to established open-source projects
- Systems and performance work in Rust, C, C++, and Assembly
- Developer tools and services in Go and Python
- Reproducible data workflows in R and Python
systems programming · performance · embedded · compilers · developer tooling · testing · scientific computing
I prefer small, reviewable changes backed by tests and aligned with each project's contribution guidelines.