FAQs

This section covers common questions developers and researchers have when getting started with Coral Protocol.

What is Coral Protocol?

Coral is a decentralised coordination protocol for AI agents. It provides a shared infrastructure layer for agent discovery, messaging, memory, and task delegation.

What’s the difference between Coral and traditional agent frameworks?

Most frameworks are single-stack tools that create agents for internal use. Coral is a protocol → it connects agents built across different frameworks, allowing them to work together through shared threads and standardised interfaces.

Is Coral a framework or a server?

Coral is neither a vertical framework nor just a messaging server. It’s a protocol layer → implemented through tools like the Coral Server and Coralizer, that enables agent interoperability and secure task execution.

Can I use Coral with agents built in LangChain or CrewAI?

Yes. Any agent can be Coralized using the Coralizer CLI as long as it complies with the Model Context Protocol (MCP).

How does memory work in Coral?

Coral defines three memory scopes:

• Private: Internal to the agent
• Thread: Shared across agents in a thread
• Session: Temporary state scoped to a task

These scopes help ensure isolation, traceability, and contextual accuracy.

Can I run Coral in production?

Coral is currently in active development. While many components are stable, the protocol is still evolving. Remote-mode support, persistent agent lifecycle, and distributed deployment models are in progress.

How do I contribute?

You can contribute via GitHub, open issues, or join discussions in our Discord:

• GitHub: Coral Server
• Discord: https://discord.gg/coralprotocol

Is Coral open source?

Yes. All core components → Coral Server, Coralizer CLI, protocol specifications are open source.

Still have questions? Reach out to the team at [e-mail] or jump into the Discord community to connect with others building on Coral.