Building workflows for AI agents shouldn't require you to maintain hundreds of lines of fragile Python or get lost in endless visual node-builders just to get specialized models talking to each other. I built Gl0wFlow to fix this. It is the runtime surface for GlowScript—a plain-English scripting language designed specifically for AI automation. Instead of writing infrastructure code, you just write what you want the agent to do: Plaintext use mcp filesystem set report to read file "report.txt" ask ai "Summarize this report" save result to "summary.txt" Why you might love it: First-Class AI Commands: ask ai, extract from message, and classify are built right into the language. Built for Speed: The runtime is written in Rust. It’s incredibly fast (sub-100ms overhead) and lets you transpile your scripts directly into native binaries. MCP Ready: Natively supports importing and exporting Model Context Protocol tools. Readable Logic: Plain-English syntax (repeat 3 times, retry 2 times) means your automations actually document themselves. Whether you are building webhook handlers, local agent networks, or complex business automations, Gl0wFlow keeps the logic front and center. Check out the repo and let me know what you think! I'll be hanging out in the comments all day to answer your questions. 🚀

GitHub

GlowScript scripting language specifically for AI automation

GlowScript scripting language specifically for AI automation