Insights, tutorials, and updates from the AgentFlow team
AgentFlow 0.3 brings a completely rearchitected runtime for multi-agent workflows. Run hundreds of concurrent agents with shared state, automatic checkpointing, and fault recovery. This release also introduces a new visual debugger for inspecting agent decisions in real time.
How we built an evaluation framework that catches agent regressions before they reach production. A deep dive into metrics, benchmarks, and automated testing.
Everything you need to know about taking your agent from local development to a scalable production environment with monitoring and observability.
Explore the different memory strategies available in AgentFlow. From short-term conversation buffers to long-term knowledge graphs, learn which pattern fits your use case.
Highlighting the best projects, contributions, and ideas from the AgentFlow community this quarter. From creative agent architectures to novel tool integrations.
When should agents act autonomously and when should they ask for permission? A framework for designing effective human-in-the-loop approval flows.
How first-class streaming support in AgentFlow lets you build responsive interfaces that show agent reasoning, tool calls, and final outputs in real time.