This is a "search engine" I built a couple of weeks ago while thinking about the agentic economy. Our agents need a way to find other agents and tools that help them complete complex tasks.
It can be used as an MCP or via A2A, your agent can find other agents and tools based on intention like "booking flight form SVQ to SFO" and get the ranked results.
It also contains filters for different protocols, auth levels, etc.
The internal ingestion tests every single public endpoint to eliminate those that are not really available, so every entity marked with "ready" can be used right away by your agents.
> Most importantly, you need to carefully engineer the learning process, so that you are not simply compiling an ever growing laundry list of assertions and traces, but a rich set of relevant learnings that carry value through time. That is the hard part of memory, and now you own that too!
I am interested in knowing more about how this part works. Most approaches I have seen focus on basic RAG pipelines or some variant of that, which don't seem practical or scalable.
Edit: and also, what about procedural memory instead of just storing facts or instructions?
It can be used as an MCP or via A2A, your agent can find other agents and tools based on intention like "booking flight form SVQ to SFO" and get the ranked results.
It also contains filters for different protocols, auth levels, etc.
The internal ingestion tests every single public endpoint to eliminate those that are not really available, so every entity marked with "ready" can be used right away by your agents.