←back to thread

What if you don't need MCP at all?

(mariozechner.at)
237 points jdkee | 4 comments | 16 Nov 25 18:58 UTC | HN request time: 0s | source
Show context
Bobaso ◴[16 Nov 25 21:01 UTC] No.45948367[source]▶
Moderne Ai agent tool have have a setting where you can trimm down the numbers of tools from an MCP server. Usefull to avoid overwhelming the LLM with 80 tools description when you only need 1
replies(2): >>45948538 #>>45949364 #
1. the_mitsuhiko ◴[16 Nov 25 21:24 UTC] No.45948538[source]▶
I don't find that to help much at all, particularly because some tools really only make sense with a bunch of other tools and then your context is already polluted. It's surprisingly hard to do this right, unless you have a single tool MCP (eg: a code/eval based tool, or an inference based tool).
replies(1): >>45948643 #
2. stavros ◴[16 Nov 25 21:37 UTC] No.45948643[source]▶
Don't you have a post about writing Python instead of using MCP? I can't see how MCP is more efficient than giving the LLM a bunch of function signatures and allow it to call them, but maybe I'm not familiar enough with MCP.
replies(1): >>45948662 #
3. the_mitsuhiko ◴[16 Nov 25 21:40 UTC] No.45948662[source]▶
> Don't you have a post about writing Python instead of using MCP?

Yes, and that works really well. I also tried various attempts of letting agents to write code that exposes MCP tool calls via an in-language API. But it's just really, really hard to work with because MCP tools are generally not in the training set, but normal APIs are.

replies(1): >>45948714 #
4. stavros ◴[16 Nov 25 21:48 UTC] No.45948714{3}[source]▶
Yeah, I've always thought that your proposal was much better. I don't know why one of the big companies hasn't released something that standardised on tool-calling via code, hm.