aosp-part-ai-and-devices

Explain AOSP AI-enabled surfaces connecting agents, apps, and companion devices.

122|30|Updated Mar 18, 2026
One-click install
npx skills add https://github.com/aospbooks/aosp-internal-book --skill aosp-part-ai-and-devices
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: aosp-part-ai-and-devices
Source: https://github.com/aospbooks/aosp-internal-book/tree/main/agents/claude/skills/aosp-part-ai-and-devices
Command: npx skills add https://github.com/aospbooks/aosp-internal-book --skill aosp-part-ai-and-devices

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This part helps readers understand how on-device AI surfaces in AOSP connect agents, apps, and companion devices for cross-device experiences.

Core Features & Use Cases

  • Understand AppFunctions as a typed, cross-application capability for invoking target app functions by assistants.
  • Explore OnDeviceIntelligence, NNAPI, Computer Control, and AppSearch-based metadata workflows to enable on-device ML inference and smart automation.
  • Examine CompanionDeviceManager and VirtualDevice frameworks to pair devices, manage presence, and render virtual displays/cameras/inputs for cross-device experiences.

Quick Start

Start by reviewing Chapter 50-ai-appfunctions.md for AppFunctions, NNAPI, and on-device ML, then Chapter 51-companion-virtual-device.md for CompanionDeviceManager and virtual devices.

Frequently Asked Questions about aosp-part-ai-and-devices

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

FAQPage Schema
How does AppFunctions in AOSP connect AI assistants to target apps?

AppFunctions in AOSP is a typed, cross-application capability that enables AI assistants to discover and invoke target app functions. It exposes app capabilities so agents can trigger internal workflows directly.

How do I use CompanionDeviceManager to manage cross-device presence and pairing?

CompanionDeviceManager in AOSP handles pairing devices and managing presence for cross-device workflows. You use it to establish secure connections and coordinate communication between companion devices.

What is the role of NNAPI in on-device ML inference within AOSP?

NNAPI enables on-device ML inference within AOSP by providing a hardware acceleration layer for machine learning models. It works with OnDeviceIntelligence to execute models directly on the device.

Can I render virtual displays and inputs using the AOSP VirtualDevice framework?

Yes, the VirtualDevice framework in AOSP allows you to render virtual displays, cameras, and inputs for cross-device experiences. It enables rendering remote device surfaces locally within your application.

Does AOSP support AppSearch-based metadata workflows for smart automation?

Yes, AOSP supports AppSearch-based metadata workflows to enable smart automation. These workflows integrate with OnDeviceIntelligence to index application content and trigger automated actions.

What are the security considerations when implementing cross-device computer control in AOSP?

Security considerations for cross-device computer control in AOSP involve securing data flows between paired devices and validating agent permissions. Architecture patterns ensure that virtual inputs and displays are isolated and managed safely.