Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Show HN: Mcptoon – Token-efficient MCP CLI client

github.com

Post reply on HN
Print view View original

21–30 of 52 posts

Re: Show HN: Mcptoon – Token-efficient MCP CLI client

#21

Post by saretup »

> zero information lost

You're just returning the name of the tool, the rest of the information (description/input schema) is definitely lost. Cut to the LLM making mistakes in calling the tool with incorrect schema or calling the wrong tools altogether, recovering, wasting tokens and cycles.

Re: Show HN: Mcptoon – Token-efficient MCP CLI client

#23

Post by cedws »

debazel wrote: post #13

Why is it replacing true/false with T/F? true/false is already 1 token in all tokenizer I've seen. Even worse is replacing null with ∅. ∅ is a special unicode symbol that takes up 2 tokens compared to the 1 token for null...

Brand new GitHub account, brand new HN account. I stay far away from projects like this these days, they can easily be malicious. GitHub needs some kind of indicator for projects authored by tenured developers with a real identity.

Re: Show HN: Mcptoon – Token-efficient MCP CLI client

#24

Post by hnlmorg »

AmazingTurtle wrote: post #16
debazel wrote: post #13

Why is it replacing true/false with T/F? true/false is already 1 token in all tokenizer I've seen. Even worse is replacing null with ∅. ∅ is a special unicode symbol that takes up 2 tokens compared to the 1 token for null...

↲ is also two tokens instead of a simple \n lmao

How is an LF two tokens? Or were you referring to the Unicode symbol?

I took their example to mean an actual LF ASCII character but now Ive read your comment, maybe I was being too charitable?

Re: Show HN: Mcptoon – Token-efficient MCP CLI client

#25

Post by spiderfarmer »

Loic wrote: post #11

I spent more than one week, as a side project, to add an MCP server to my Cheméo website. Only 4 tools. It took me way more time than expected, I was thinking: "Just wrap the REST API, 2h, done". The MCP payload has nothing to do with the REST API one. Because you need to make it interpretable and context efficient even so it is structured data. It was really interesting work and I suppose very little people are taki…

This is where using a framework really shines. I used Laravel MCP which makes it trivial to add MCP tools to your CRUD.

Re: Show HN: Mcptoon – Token-efficient MCP CLI client

#26

Post by maxrev17 »

Loic wrote: post #11

I spent more than one week, as a side project, to add an MCP server to my Cheméo website. Only 4 tools. It took me way more time than expected, I was thinking: "Just wrap the REST API, 2h, done". The MCP payload has nothing to do with the REST API one. Because you need to make it interpretable and context efficient even so it is structured data. It was really interesting work and I suppose very little people are taki…

Yeah this is why a code execution sandbox so the ai can batch calls and select from the response format what it wants and limit the number of responses with instruction to be concise and preserve its context is a really cool thing to do.

Re: Show HN: Mcptoon – Token-efficient MCP CLI client

#28

Post by fallingbananna »

debazel wrote: post #13

Why is it replacing true/false with T/F? true/false is already 1 token in all tokenizer I've seen. Even worse is replacing null with ∅. ∅ is a special unicode symbol that takes up 2 tokens compared to the 1 token for null...

What's even weirder is that the substitution is explained in "How TOON works" section. Yet TOON never describes such behavior anywhere in its spec.

Re: Show HN: Mcptoon – Token-efficient MCP CLI client

#29

Post by eterm »

moinism wrote: post #22

How do unresearched, vibe-coded projects like this reach the front page?

I'm convinced that the majority of upvotes are based on reading a title rather than clicking through to an article.

People want a token efficient MCP CLI client. Whether this actually is one is less relevant.

Re: Show HN: Mcptoon – Token-efficient MCP CLI client

#30

Post by denis-stable »

bobkinartem wrote: post #10

I thought Codex and Claude Code agents are already token-efficient so writing agents that saves tokens is pointless.

I don't know about Codex, but Claude Code defers loading tools if their definitions exceed 10% of the context, https://code.claude.com/docs/en/mcp#how-it-works.
Post reply on HN