Magisterium AI Now Supports the A2A Protocol

For the past several months, developers have been connecting their AI tools to Magisterium AI through our MCP server — bringing the full depth of the Catholic intellectual tradition into tools like Claude, ChatGPT, and custom AI workflows.

Today, we're taking the next step: Magisterium AI now supports the A2A (Agent-to-Agent) protocol, an open standard that lets AI agents discover, communicate with, and delegate tasks to each other.

MCP vs. A2A — What's the Difference?

If you've used our MCP server, you already know the value of connecting AI tools to authoritative Catholic sources. MCP and A2A are complementary:

• MCP lets AI tools access Magisterium's knowledge. You add our server to Claude or ChatGPT, and the tool can search documents, look up readings, or ask theological questions on your behalf.
• A2A lets AI agents collaborate with Magisterium as a peer. An orchestrating agent — built on LangChain, CrewAI, or any framework — can discover Magisterium's capabilities automatically, send it structured tasks, and receive cited results through a standard JSON-RPC interface.

Think of MCP as plugging Magisterium into your AI toolbox. A2A is more like adding Magisterium to your AI team.

What Can A2A Do?

Magisterium AI exposes five skills through the A2A protocol, each mapped to the same knowledge infrastructure that powers magisterium.com:

Catholic Q&A

Ask any theological question and receive a comprehensive, cited answer drawn from 29,000+ Magisterial documents, Scripture, and the Church Fathers.

Document Search

Search semantically across papal encyclicals, council decrees, catechism sections, and canon law.

Document Retrieval

Retrieve the full text and metadata of a specific document by ID.

Liturgical Readings

Get the Catholic Mass readings for any date — supports natural language queries like "next Sunday" or "Easter 2026."

Saints of the Day

Look up the saints commemorated on a given date from the Roman Martyrology.

Every response follows the A2A specification: structured artifacts with typed parts (text, data), full message history, and a standard task lifecycle. For technical details on discovery, authentication, and making requests, see the A2A developer docs.

Real-World Use Cases

A2A opens up scenarios that go beyond what a single chatbot can do:

• Catholic education platforms can delegate theological questions to Magisterium within their own AI study tools, presenting sourced answers inline without building their own doctrinal knowledge base.
• Parish apps and chatbots can answer questions about sacraments, marriage prep, or Church teaching by routing to Magisterium behind the scenes, while handling parish-specific queries locally.
• Multi-agent orchestrators — built on frameworks like LangChain or CrewAI — can include Magisterium as a "Catholic theology expert" in their agent roster, automatically consulting it when moral or doctrinal questions arise.
• Daily devotional apps can call liturgical_readings and saints_of_the_day every morning to populate their content programmatically.

In every case, the end user interacts with their app. Magisterium works in the background, providing the authoritative knowledge layer.

Pricing and Rate Limits

A2A is available on the Pro, Organization, and Enterprise plans. It shares the same rate-limit pool as MCP — a request via A2A counts against the same quota as a request via MCP. Free accounts are not able to access the A2A endpoint.

• Pro
  15 Requests / Minute
  2 Q&A Requests / Minute
• Organization
  15 Requests / Minute
  2 Q&A Requests / Minute
• Enterprise
  Custom rate limit

For full pricing details, see the MCP & A2A Pricing page.

Getting Started

1. Create an account at magisterium.com and ensure you're on a paid plan.
2. Fetch the Agent Card at https://www.magisterium.com/.well-known/agent.json to see available skills.
3. Authenticate using the same OAuth flow as MCP.
4. Send your first message/send request — see the Getting Started guide for curl, Python, and TypeScript examples.

Full documentation is available in our A2A developer docs.

What's Next

This is our initial A2A release. On the roadmap:

• API key authentication for simpler server-to-server integrations
• SSE streaming for real-time responses from the Catholic Q&A skill
• Multi-turn conversations with context carried across related tasks
• Agent registry listings for broader discovery

We're excited to see what you build. If you have questions or want to share what you're working on, reach out to us.