Coding Agents Workshop — Development skill for Claude Code

Optimizing Codebases for Agents

[](https://blog.sshh.io/) [](https://github.com/sshh12) [](https://x.com/ShrivuShankar) [](https://linkedin.com/in/shrivushankar) [](https://sshh.io/coffee-chat)

**Conference:** Coding Agents: AI Driven Dev Conference **Date:** March 3, 2026 **Location:** Computer History Museum, Mountain View, CA **Speaker:** Shrivu Shankar (VP AI, Abnormal Security)

Your agent isn't broken. Your codebase is.

What This Is

This repo contains the materials for the "Optimizing Codebases for Agents" workshop. It includes a demo app in two versions (before and after optimization), an AI-readiness scorecard, and agent race narrator notes.

Prerequisites

• A laptop with a development environment
• An AI coding tool installed and authenticated (Claude Code, Gemini CLI, Codex CLI, or Cursor)
• Python 3.10+
• A repository you want to audit (open source or personal). If you don't bring one, you can audit the A/ demo app.

Quick Start

git clone https://github.com/sshh12/coding-agents-workshop.git
cd coding-agents-workshop

Run the "Before" app (Version A)

cd A
pip install -r requirements.txt
python app.py
# Open http://localhost:8000

Run the "After" app (Version B)

cd B
pip install -r requirements.txt
python manage.py run
# Open http://localhost:8000

Run tests (Version B only)

cd B
pytest

Repo Structure

coding-agents-workshop/
  README.md           # You are here
  scorecard.md        # AI-Readiness Audit (use