skills.rest
digitalocean-labs
digitalocean-labsOFFICIAL
Explore all skills by this author →
WebsiteGitHub

ai-services

Configure Gradient AI serverless inference and ADK agents for App Platform using Python 3.13 and doctl.

35|4|Updated Dec 28, 2025
Search Tags:#ai#agent#adk#app-platform#gradient-ai#serverless-inference
One-click install
npx skills add https://github.com/digitalocean-labs/do-app-platform-skills --skill ai-services
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ai-services
Source: https://github.com/digitalocean-labs/do-app-platform-skills/tree/main/skills/ai-services
Command: npx skills add https://github.com/digitalocean-labs/do-app-platform-skills --skill ai-services

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Configure DigitalOcean Gradient AI Platform for App Platform applications to enable both serverless inference and full AI agents with the ADK, simplifying AI integration and deployment.

Core Features & Use Cases

  • Serverless Inference for OpenAI-compatible APIs to run models without managing agents.
  • Agent Development Kit (ADK) to build, deploy, and manage AI agents with knowledge bases, RAG, guardrails, and multi-agent workflows.
  • Credential handling and integration: manage model access keys via GitHub Secrets or App Platform Secrets, and wire them into app specs.
  • Quick Start guides and reference docs to accelerate adoption.

Quick Start

Add serverless inference endpoints and an ADK-enabled agent to your App Platform project to start building AI capabilities.

Frequently Asked Questions about ai-services

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

FAQPage Schema
How do I configure serverless inference for an App Platform application?▼

Configure serverless inference for an App Platform application by adding inference endpoints and wiring model access keys into your app specs via App Platform Secrets. This enables OpenAI-compatible APIs to run models without managing agents.

Can I build AI agents with knowledge bases and guardrails on App Platform?▼

Build AI agents with knowledge bases, guardrails, and multi-agent workflows on App Platform by using the Agent Development Kit. The ADK provides fully managed agents to simplify AI integration and deployment.

Do I need Python 3.13 to deploy AI services on App Platform?▼

Python 3.13 is required to deploy AI services on App Platform using the Agent Development Kit. You must also install the gradient-adk package and use doctl-based deployment workflows for the application.

What is the best way to manage model access keys for Gradient AI services?▼

Manage model access keys for Gradient AI services by storing them securely in GitHub Secrets or App Platform Secrets. This approach safely wires credentials into your app specs without hardcoding sensitive data.

Does the Agent Development Kit support RAG and multi-agent workflows?▼

The Agent Development Kit supports RAG, guardrails, and multi-agent workflows for AI services. It allows you to build, deploy, and manage fully managed agents with knowledge bases on App Platform.

Why use serverless inference instead of deploying a full AI agent?▼

Use serverless inference instead of a full AI agent when you need OpenAI-compatible APIs to run models without the overhead of managing agents. It provides a simpler integration for applications requiring direct model access.

Related Skills

Explore More Software Engineering Skills→
ClyptAIClyptAI

ai-services

Configure Gradient AI serverless inference and the Agent Development Kit for App Platform deployments.

Official
Advanced
litestar-orglitestar-org
13

litestar-ai-serving

Build HTTP AI agent endpoints with Litestar and Google ADK.

Official
Advanced
digitalocean-labsdigitalocean-labs
35

postgres

Configure DigitalOcean Postgres with per-app schemas and bindable credentials.

Official
Advanced
peteromalletpeteromallet
97

serving-llms-vllm

Deploy LLM inference with vLLM using OpenAI-compatible endpoints, quantization, and tensor parallelism.

Community
Intermediate