December 16, 2025
Stop prompting, start conducting
Show HN: Zenflow – orchestrate coding agents without "you're right" loops
Hackers cheer as Zenflow corrals AI “slop” and lets models tag‑team
TLDR: Zenflow promises to wrangle multiple AI coding agents by following a spec and auto‑verifying results. Commenters cheered the “no slop” stance and multi‑model support, debated prompts vs orchestration, and flagged a Firefox download glitch—making it both hype and hiccup for teams eyeing AI‑powered development.
Zenflow just crashed the party with an orchestration platform that claims to keep AI coding agents from those cringe “you’re right” loops—and the crowd went loud. Fans say it’s like hiring a project manager for your robot helpers: it reads your spec, runs tasks in parallel, and auto‑checks the work before you see it. One commenter even called it the next step after spec‑driven development, praising the bring‑your‑own‑bot setup (Claude, Gemini, etc.).
But the real spice? The site’s bold stance against bad AI output—calling it “slop.” The community loved the honesty. “Finally, a team that says the quiet part out loud,” cheered one user, while memes flew of AI “slop soup” getting poured down the drain. The CEO’s hot take—“AI doesn’t need better prompts. It needs orchestration.”—sparked debate, with prompt true believers side‑eyeing the idea while pragmatic devs shouted “preach.”
Hands‑on testers reported early wins: fewer hallucinations (AI making things up), better oversight, and flexible workflows that don’t lock you in. Skeptics still want proof it won’t fall apart when projects take weird turns. And of course, tech drama: a Firefox user flagged that advanced tracking protection breaks the download form. Cue jokes about “orchestration, but first disable your shields.”
If you want to watch the AI agents behave, the crowd says try Zenflow—just maybe don’t use super‑protective browser settings.
Key Points
- •Zenflow is a multi-agent orchestration platform focused on reliable AI-driven software development.
- •It uses spec-driven workflows so agents align to PRDs/specs, reducing drift and scope creep.
- •Parallel execution runs tasks in isolated environments, coordinated within each workflow.
- •Built-in verification includes automated tests, cross-agent review, and quality gates before human review.
- •The site offers an app download, notes a Windows version is coming soon, and provides a Slack community and documentation.