adcp-si

Coordinate AdCP SI session lifecycle events with consent handling and status updates.

1|5|Updated Nov 5, 2025
One-click install
npx skills add https://github.com/adcontextprotocol/adcp-client-python --skill adcp-si
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: adcp-si
Source: https://github.com/adcontextprotocol/adcp-client-python/tree/main/skills/adcp-si
Command: npx skills add https://github.com/adcontextprotocol/adcp-client-python --skill adcp-si

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

AdCP Sponsored Intelligence (SI) session orchestration for brand-agent conversations, enabling seamless setup, consent-aware identity handling, offering previews, messaging, and orderly termination in a single, programmable flow.

Core Features & Use Cases

  • Coordinates the four SI tasks si_initiate_session, si_send_message, si_get_offering, and si_terminate_session to manage a complete conversational lifecycle.
  • Handles consent models to switch between anonymous and personalized sessions and maintains session continuity with tokens.
  • Supports offering previews and structured messaging flows to enable conversational commerce with brand agents across channels.

Quick Start

Initiate a session against a brand agent using si_initiate_session, then drive the conversation with si_send_message and si_terminate_session as needed.

Frequently Asked Questions about adcp-si

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

FAQPage Schema
How do I orchestrate conversational commerce sessions between buyers and brand agents?

You orchestrate conversational commerce sessions by initiating a session, sending messages, previewing offerings, and terminating the conversation across channels. This coordinates the complete AdCP Sponsored Intelligence lifecycle with consent handling and status updates.

What is AdCP Sponsored Intelligence session management for brand-agent conversations?

AdCP Sponsored Intelligence session management is a protocol for coordinating lifecycle events between buyers and brand agents. It enables consent-aware identity handling, offering previews, structured messaging, and orderly termination in a single programmable flow.

How do I handle identity consent when initiating conversational commerce across channels?

Handle identity consent by applying consent models that switch between anonymous and personalized sessions. The orchestration maintains session continuity using tokens while aligning with the AdCP SI protocol for status updates and robust error handling.

Can I preview brand offerings during an active messaging flow?

Yes, you can preview brand offerings during an active messaging flow. The orchestration supports offering retrieval alongside structured messaging, allowing buyers to interact with brand agents and preview offerings throughout the active session.

What is the best way to terminate a brand-agent conversation session without losing state?

The best way to terminate a brand-agent conversation is by executing the session termination task within the orchestration flow. This ensures orderly conversation closure, aligns with the AdCP SI protocol, and handles any remaining lifecycle events or errors properly.

Does conversational commerce session orchestration work without external dependencies?

Yes, conversational commerce session orchestration works without external dependencies. It independently coordinates the four SI tasks for initiating, messaging, offering retrieval, and terminating sessions while managing consent and error handling internally.