nuvel

Scaffold Google ADK agent projects with workflow graphs and ACP compatibility.

2|1|Updated Apr 4, 2026
One-click install
npx skills add https://github.com/Folken2/nuvel --skill nuvel
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: nuvel
Source: https://github.com/Folken2/nuvel/tree/main/.claude/skills/nuvel
Command: npx skills add https://github.com/Folken2/nuvel --skill nuvel

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill solves the complexity of architecting and scaffolding production-ready Google ADK agents by providing a standardized, battle-tested framework and a library of canonical knowledge.

Core Features & Use Cases

  • Agent Scaffolding: Rapidly generate structured ADK project skeletons with optional support for personas, Composio tool routing, workflow graphs, and Agent Client Protocol (ACP) compatibility.
  • Knowledge Integration: Access a curated library of ADK patterns, including prompt engineering, tool creation, human-in-the-loop callbacks, and streaming.
  • Use Case: Use this skill to build a complex, multi-step agent that requires a planning-execution workflow, or to bootstrap a personal agent that evolves its own skills over time.

Quick Start

Use the nuvel skill to scaffold a new Google ADK agent named customer-support-bot with a description of an agent that handles incoming support tickets.

Frequently Asked Questions about nuvel

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

FAQPage Schema
How do I scaffold a production-ready Google ADK agent?

Scaffolding a Google ADK agent involves using a meta-agent CLI to generate structured project skeletons with optional support for personas, workflow graphs, and ACP editor integrations.

What architectural patterns can I use when building ADK agents?

When building ADK agents, you can implement diverse architectural patterns including workflow graphs for multi-step planning, self-evolving personas, and ACP-compliant editor integrations.

Do I need a Python environment to use the nuvel meta-agent CLI?

Yes, using the meta-agent CLI requires a Python environment with the nuvel package installed to manage agent skeletons and provide domain-specific ADK knowledge.

What's the best way to bootstrap an agent that evolves its own skills over time?

To bootstrap an agent that evolves its own skills over time, use a scaffolding meta-agent that supports self-evolving personas and integrates a curated library of ADK patterns.

Does ADK agent scaffolding support human-in-the-loop callbacks and streaming?

Yes, ADK agent scaffolding supports these features by providing access to a curated knowledge library of ADK patterns covering human-in-the-loop callbacks, streaming, and tool creation.

When should I not use a meta-agent for scaffolding ADK workflows?

Avoid using a meta-agent for scaffolding ADK workflows if your project requires highly custom, non-standard architectural patterns outside typical workflow graphs or ACP integrations.