[Feature Request] Add GibsonAI "Memori" memory integration

[Astrodevil]

Required prerequisites

• [x] I have searched the Issue Tracker and Discussions that this hasn't already been reported. (+1 or comment there if it has.)
• [ ] Consider asking first in a Discussion.

Motivation

Request to add a new Camel integration with an open-source AI memory layer, "Memori," for multi-agent memory management.

Solution

Memori comes with database support like SQLite, PostgreSQL, MySQL. It also has dual memory mode for better multi-agent usages.

Alternatives

No response

Additional context

We have already added Camel integration with Memori

• Official docs
• Code example

[OmCheeLin]

We still need to discuss this issue and clarify some things. We currently have three implementations of Memory, ChatHistoryMemory、VectorDBMemory、LongtermAgentMemory If we want to integrate Memori, which level of the existing memory system should it be used as? Inherit AgentMemory or only use it as storage? This is the official document. Code example basic usage Memori seems to have complete functions, but I have some doubts about how to integrate it into camel AI.

[OmCheeLin]

@Wendong-Fan cc

[Wendong-Fan]

thanks @OmCheeLin and @Astrodevil , after double-checking the example code (https://github.com/GibsonAI/memori/blob/main/examples/integrations/camelai_example.py), I'm still wondering how Memori integrates with the CAMEL agent's memory. It looks like it might still be using CAMEL's built-in agent memory.

@Astrodevil, could you please clarify how this integration works?

[Astrodevil]

I will check and revert.