April 5, 2026
Specs, bots, and a name fight
Show HN: Modo – I built an open-source alternative to Kiro, Cursor, and Windsurf
He built an open‑source AI coder in days—fans hype, skeptics want receipts, name triggers déjà vu
TLDR: An indie dev launched Modo, a free, open‑source AI coding app that plans work before writing code. The crowd loves the built‑in structure and editor integration, but debates whether it beats simple scripts, begs for “session replay” to verify AI moves, and laughs about the name already being taken.
An indie dev just dropped “Modo,” an open‑source AI coding app that actually plans before it writes code—think: it makes a mini blueprint, to‑do list, and project rules, then gets to work. The tool bundles neat tricks like a plan‑first workflow, a Run Task button for your to‑dos, always‑on project rules, and a flip between hands‑off Autopilot and supervised mode. It’s basically a DIY answer to paid AI code editors like Cursor or Windsurf.
But the comments stole the show. One crowd is pumped about baking this agent brain right into the editor—multiple threads, sub‑agents, even “powers” for React and Docker—and they’re already asking for receipts: a session replay to watch what the AI did, like a crime scene time‑lapse for your code. Another camp is politely skeptical: “Cool, but how is this better than a few files and prompts?” They want proof that Modo’s structure beats simple scripts or a markdown checklist. And then came the spice: the name fight. Veterans pointed out “Modo” is already a famous 3D app, sparking friendly groans and “rename when?” jokes.
Overall vibe: ambitious, hackable, and a little cocky—the dev says hitting 60–70% of the big apps wasn’t that hard. Fans cheer. Skeptics say: show us the receipts, and maybe change the name.
Key Points
- •Modo is an open-source, MIT-licensed AI IDE that emphasizes planning before coding.
- •Built on the Void editor (a VS Code fork), it adds features beyond AI chat, inline edits, autocomplete, LLMs, and MCP integration.
- •Spec-Driven Development organizes work into requirements, design, and tasks stored in .modo/specs/, with persistent tasks and an executable task list.
- •Steering Files and Agent Hooks enable project-wide rules and automated actions across 10 event types, including command execution and agent interactions.
- •Additional features include Autopilot/Supervised mode, parallel chat sessions, Vibe/Spec modes, Subagents, installable Powers (e.g., TypeScript, React, Docker), and a Dedicated Explorer Pane.