skills.rest
Folken2
Folken2COMMUNITY
Explore all skills by this author →
WebsiteGitHub

adk-workflow-graphs

Orchestrate agentic workflows with conditional routing and parallel execution using google-adk.

2|1|Updated Apr 4, 2026
Search Tags:#automation#workflow#agents#orchestration#adk#graph
One-click install
npx skills add https://github.com/Folken2/nuvel --skill adk-workflow-graphs
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: adk-workflow-graphs
Source: https://github.com/Folken2/nuvel/tree/main/nuvel/backends/adk/skills/adk-workflow-graphs
Command: npx skills add https://github.com/Folken2/nuvel --skill adk-workflow-graphs

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires google-adk, and includes references (resource) components.

What problem does it solve?

This skill solves the challenge of orchestrating complex agent logic that goes beyond simple linear or parallel execution, allowing for conditional branching, revision loops, and dynamic task routing.

Core Features & Use Cases

  • Graph-Based Orchestration: Define nodes and edges to create sophisticated agent pipelines with conditional logic and fan-out/fan-in patterns.
  • Human-in-the-Loop Cycles: Implement robust revision loops where agents can pause for human feedback before proceeding.
  • Dynamic Node Execution: Execute child nodes at runtime based on data-driven requirements, enabling recursive or adaptive agent behavior.

Quick Start

Load the adk-workflow-graphs skill to define a new workflow with conditional routing and parallel execution nodes.

Frequently Asked Questions about adk-workflow-graphs

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I orchestrate complex agent workflows with conditional routing and loops?▼

Graph-based orchestration organizes agent pipelines into nodes and edges to execute conditional branching, parallel fan-out, and revision loops. It is needed when your multi-step task automation requires execution paths that depend on runtime data or iterative refinement.

How does human-in-the-loop workflow execution work for agent revision cycles?▼

To build graph-based agent pipelines, load the skill and define nodes for specific tasks alongside edges for conditional routing. You can configure parallel execution and fan-out patterns to orchestrate complex multi-step automation dynamically.

Do I need google-adk to manage node state and edge transitions in workflows?▼

Yes, this skill requires the google-adk library version 2.0.0 or higher. It relies on google-adk specifically to manage node state and edge transitions during complex agent orchestration and dynamic node execution.

What's the best way to automate multi-step tasks with dynamic node execution?▼

The best way to automate multi-step tasks is using graph-based workflows that execute child nodes dynamically at runtime based on data-driven requirements. This approach handles conditional branching and recursive agent behavior effectively.

Related Skills

Explore More Software Engineering Skills→
kevinnguyen271090kevinnguyen271090
15

google-adk-python

Build and deploy AI agents with Google ADK Python.

Community
Basic
zxc1a1a1zxc1a1a1

adk-runtime-contract

Standardizes Child Agent ADK runtime contracts for configuration, streaming, artifacts, permissions, error handling, and observability.

Community
Advanced
distribucionesfayosmoreno-sysdistribucionesfayosmoreno-sys

workflow-orchestration-patterns

Orchestrate distributed system workflows using Temporal saga, entity, and fan-out/fan-in patterns.

Community
Advanced
codejunkie99codejunkie99
101

gf-build

Orchestrate parallel SystemVerilog builds with phase-based decomposition and integration.

Community
Advanced