---
title: "Using Agents to Build Complex Apps That Work"
url: https://stacklist.com/card/9fda4d09-e267-418f-95ff-972a7c87069e
source_url: "https://websummercamp.com/2026/session/0-140k-loc-using-agents-to-build-complex-apps-that-actually-work"
stack: https://stacklist.com/stack/3d40319c-c76c-4ae5-b09d-40eea149cbc9
summary: "Using Agents to Build Functional Complex Apps discusses how to ship complex software entirely with AI agents, using SpriteMachine (a professional pixel-art editor with 140,000 lines of TypeScript) as a case study. The key insight is that agentic coding success depends on architecture and enforcement mechanisms that allow codebases to validate agent output, not on prompting techniques alone."
tags: "agentic-coding, agent-architecture, complex-apps, software-engineering, code-generation, typescript, webgpu"
key_entities: "Hrvoje Čop (person), Polymachine (organization), SpriteMachine (technology), WebGPU (technology), TypeScript (technology), agentic-coding (concept), code-generation-with-agents (concept)"
classification: "transcript"
content_hash: "sha256:e498d8bc279528247f97aae8d5cbbd31e9eaafd37cdec038aae567d5c965f772"
acp_version: "0.2"
token_counts_approximate: 430
visibility: public
agent_accessible: true
status: "final"
---

# Using Agents to Build Complex Apps That Work

Description Everyone has the "agentic coding best practices" talk: prompt like this, use subagents, write a good context file. But look at who's actually shipping complex software entirely with agents — it's either enterprise teams with armies, or it's todo-app demos. The interesting middle is mostly empty. This talk is a field report from that middle. SpriteMachine is a professional pixel-art editor — a custom WebGPU renderer, real-time pointer and pressure pipeline, deterministic undo/redo, a sandboxed scripting engine, dual web/desktop deployment. ~140,000 lines of TypeScript and 45 hand-tuned GPU shaders. Not one line of source was written by hand. I'll argue the bottleneck on agentic coding isn't the model — it's whether your codebase can tell the agent it's wrong before you have to. We'll walk through the concrete machinery that makes an all-agent codebase hold together instead of rotting. None of it is AI-specific technology. What's new is the reason: your fastest contributor now has zero judgment, so legibility and enforcement stop being nice-to-haves. Build a codebase that argues back. Key takeaway Shipping complex apps with agents is an architecture-and-enforcement problem, not a prompting problem — and you can apply all of it in any stack on Monday. Speaker CTO @ Polymachine Hrvoje Čop is co-founder and CTO at Polymachine, where he spends his days automating things that weren't supposed to be automated and inventing ways for a tiny company to do things big ones often can't. With 25 years in 3D graphics and programming — and a healthy disregard for doing things the normal way — he's been spotted speaking at conferences and herding render nodes