Skip to content
View aelaguiz's full-sized avatar
🤖
Building Doctrine + Rally
🤖
Building Doctrine + Rally
39 followers · 14 following

Achievements

Achievement: Pair Extraordinairex3Achievement: YOLOAchievement: QuickdrawAchievement: Pull Sharkx2Achievement: Arctic Code Vault Contributor

Achievements

Achievement: Pair Extraordinairex3Achievement: YOLOAchievement: QuickdrawAchievement: Pull Sharkx2Achievement: Arctic Code Vault Contributor

Highlights

Organizations

@cratejoy

Block or report aelaguiz

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.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. 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
aelaguiz/README.md

Hi, I'm Amir 👋

📍 Austin, TX | 🤖 Building agent infrastructure | 🏁 Doctrine + Rally | 🚀 YC S13 founder @ Poker Skill + Cratejoy

Python TypeScript AI Agents MCP Codex Claude CLI GitHub

I build the boring but necessary layer that makes agents usable in real work: typed instructions, visible state, recoverability, and operator control. The goal is simple: make agent workflows feel like software instead of prompt soup.

Current Projects

  • 📜 Doctrine — typed DSL and compiler for reusable AI agent instructions and workflows that compile to AGENTS.md
  • 🏁 Rally — filesystem-first harness for repo-local multi-agent coding workflows
  • 🧪 codex-autoresearcher — worker/judge harness for optimizing anything you can measure
  • 🧰 aimgr — operator tooling for AI accounts, browsers, and local agent rails
  • 🌐 mcp-url-fetch — MCP server for fetching URLs from Claude, Cline, and other clients
  • 🔔 codex-notifier — audible notifications for Codex CLI on macOS
  • 🧠 langdspy — earlier public line of work on declarative prompting and optimization

Why this stack

  • Doctrine is the authoring and compiler layer.
  • Rally is the runtime and harness layer.
  • Together they aim to make AI workflows typed, inspectable, and recoverable.

Background

  • Founder / CEO: Poker Skill, Cratejoy
  • YC S13
  • Previously: MarketZero (acquired by Zynga)
  • Older ML / open-source lineage: pyvotune, Theano, DeepLearningTutorials

Start here

  • If you want the authoring and compiler layer, start with Doctrine.
  • If you want the runtime and harness layer, start with Rally.
  • If you want the earlier public line into this work, start with langdspy.
Selected earlier public work

Pinned Loading

  1. doctrine doctrine Public

    Code-like DSL and compiler for agent workflows that compile to portable AGENTS.md instructions.

    Python 1 1

  2. rally rally Public

    Filesystem-first agent harness for strong, stable coding-agent workflows with repo-local run state.

    Python

  3. langdspy langdspy Public

    Declarative prompting and optimization for LangChain-style pipelines. Enterprise ready.

    Python 7 3

  4. codex-autoresearcher codex-autoresearcher Public

    Codex-native autoresearch harness with structured worker/judge turns for optimizing anything you can measure.

    Python 3

  5. aimgr aimgr Public

    AI account and browser manager for Codex, Claude, Pi, and OpenClaw operator workflows.

    JavaScript 1

  6. mcp-url-fetch mcp-url-fetch Public

    Model Context Protocol server for fetching URLs from Claude, Cline, and other MCP clients.

    Python 2