Raph Terrier

@joeybaker 's "Should agents be durable?" workshop at @AICouncilConf was packed. We demoed self-orchestrating agents that don’t require pre-provisioned infrastructure, custom retry logic, or separate configuration to manage state or recovery.

Ask a research question out loud. Under 60 seconds later, you have a complete, sourced answer. We built a reference architecture with @Render using AssemblyAI's Voice Agent API + Render's new Workflows. Core insight: keep the voice channel separate from background orchestration. Don't block audio waiting on tool execution. The stack: → @AssemblyAI Voice Agent API handles real-time audio streaming → @render Workflows classify, plan, search, synthesize as isolated tasks → @mastra agents classify question "shapes" before searching → @youdotcom powers parallel search branches Repo includes the Render Blueprint, Mastra configs, and a live demo. Full tutorial + source code: assemblyai.com/blog/voice-age…

Deploy headless Flue agents on Render in one click. Template ships with two example agents to get you started. To add a new one, drop a .ts file in .flue/agents/ and push. render.com/templates/flue

Introducing Flue — The First Agent Harness Framework Flue is a TypeScript framework for building the next generation of agents, designed around a built-in agent harness. Flue is like Claude Code, but 100% headless and programmable. There's no baked in assumption like requiring a human operator to function. No TUI. No GUI. Just TypeScript. But using Flue feels like using Claude Code. The agents you build act autonomously to solve problems and complete tasks. They require very little code to run. Most of the "logic" lives in Markdown: skills and context and AGENTS.md. Flue is like Astro or Next.js for agents (not surprising, given my background 🙃). It's not another AI SDK. It's a proper runtime-agnostic framework. Write once, build, and deploy your agents anywhere (Node.js, Cloudflare, GitHub Actions, GitLab CI/CD, etc). We originally built Flue to power AI workflows inside of the Astro GitHub repo. But then @_bgiori got his hands on it, and we realized that every agent needs a framework like Flue, not just us. Check it out! It's early, but I'm curious to hear what people think. Are agents ready for their library -> framework moment?

Building reliable apps and agents shouldn’t mean managing queues, workers, and retry infrastructure. With Render Workflows, you just push your code. Turn functions into durable, composable tasks → Trigger tasks from anywhere → Execution is automatic ↓ render.com/blog/durabilit…