Agent DailyAgent Daily
releaseintermediate

[Release] google/adk-python v2.0.0b1: v2.0.0b1

By sasha-gitggithub
View original on github

Google ADK Python v2.0.0-beta.1 marks the transition to Beta phase with significant architectural improvements. Key additions include a full Workflow graph orchestration system with NodeRunner for execution isolation, explicit ReAct loop nodes for agent execution, and Human-in-the-loop resumption capabilities. The release also addresses a critical RCE vulnerability in YAML configuration handling and optimizes performance for single-turn LLM agents.

Key Points

  • Transitioned to Beta phase with updated documentation reflecting project maturity
  • Fixed critical RCE vulnerability in nested YAML configuration parsing
  • Implemented full Workflow(BaseNode) graph orchestration for complex agent workflows
  • Introduced NodeRunner for per-node execution isolation and independent node management
  • Added DefaultNodeScheduler enabling standalone node resume via ctx.run_node()
  • Replaced legacy single agent flows with explicit ReAct loop nodes for clearer agent execution patterns
  • Enabled BaseNode as root in both Runner and CLI for flexible workflow entry points
  • Implemented Human-in-the-loop (HITL) resume via event reconstruction for interactive workflows
  • Added lazy scan deduplication and resume for dynamic nodes to handle variable workflow structures
  • Optimized performance by bypassing Mesh for leaf single-turn LlmAgent instances

Found this useful? Add it to a playbook for a step-by-step implementation guide.

Workflow Diagram

Start Process
Step A
Step B
Step C
Complete
Quality

Concepts