Skip to content
View antonygiomarxdev's full-sized avatar
🏠
Working from home
🏠
Working from home

Block or report antonygiomarxdev

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
antonygiomarxdev/README.md

Antony Giomar Hernández

Product engineer in Managua, Nicaragua. Co-founder of RUMAI, reproductive intelligence for cattle herds. I build products with AI agents, and the tools that check their work.

antonygiomarx.dev · X · LinkedIn

How I build

AI agents write most of the code. I decide what gets built, set the checks (types, lint, tests and mutation testing) and answer for what ships. Agents build in parallel from a shared task list, and a reviewer that never wrote the code reads the result.

Projects

These tools came out of that way of working.

  • reapr: Mutation testing for AI agents, and for every pull request.
  • puppeteering: A field guide to running a team of Claude Code agents on real work.
  • gauntlet: A code-quality ratchet an agent cannot argue with.
  • mill: Turns one AI session into a coordinator that dispatches specialised workers.
  • maverick: Offline-first LoRaWAN gateway and network server in a single binary.
  • angular-doctor: Scans an Angular codebase and gives it a 0 to 100 health score.
  • synapse: Distributed inference for Mixture-of-Experts models on consumer GPUs.

Background

Since 2019 I have run production messaging and AI agents on the WhatsApp Business Platform. Self-taught. Meta Certified in WhatsApp for Business technical implementation.

Pinned Loading

  1. angular-doctor angular-doctor Public

    Scan your Angular codebase for performance, architecture, and dead code issues — outputs a 0–100 health score with actionable diagnostics

    TypeScript 25 1

  2. synapse synapse Public

    Distributed inference infrastructure for Mixture-of-Experts models. Run large MoE models on consumer GPUs.

    Rust 12 1

  3. maverick maverick Public

    Offline-first LoRaWAN gateway + network server in a single binary. Runs on Raspberry Pi. No cloud required.

    Rust 11

  4. mill mill Public

    Turns one AI session into a coordinator that dispatches specialised workers, with policy and gate scripts in plain files

    Shell 2 1

  5. puppeteering puppeteering Public

    A field guide to running a team of Claude Code agents on real work: one lead, a shared pile of tasks, independent review, every claim backed by a check.

    JavaScript 2

  6. gauntlet gauntlet Public

    A code-quality ratchet your agent can't argue with.

    Shell 1