Ask HN: How do you know if AI agents will choose your tool?

#1

Post by dmpyatyi » Mon, Feb 23, 2026, 7:32 PM UTC

YC recently put out a video about the agent economy - the idea that agents are becoming autonomous economic actors, choosing tools and services without human input.

It got me thinking: how do you actually optimize for agent discovery? With humans you can do SEO, copywriting, word of mouth. But an agent just looks at available tools in context and picks one based on the description, schema, examples.

Has anyone experimented with this? Does better documentation measurably increase how often agents call your tool? Does the wording of your tool description matter across different models (ZLM vs Claude vs Gemini)?

Re: Ask HN: How do you know if AI agents will choose your tool?

#2

Post by jackfranklyn » Mon, Feb 23, 2026, 7:33 PM UTC

[flagged]

Re: Ask HN: How do you know if AI agents will choose your tool?

#3

Post by JacobArthurs » Mon, Feb 23, 2026, 7:36 PM UTC

Tool description quality matters way more than people expect. In my experience with MCP servers, the biggest win is specificity about when not to use the tool. Agents pick confidently when there's a clear boundary, not a vague capability statement.

Re: Ask HN: How do you know if AI agents will choose your tool?

#4

Post by zahlman » Mon, Feb 23, 2026, 7:43 PM UTC

jackfranklyn wrote: post #2

[flagged]

> inline examples in the description beat external documentation every time. The agent won't browse to your docs page.

That seems... surprising, and if necessary something that could easily be corrected on the harness side.

> The schema side matters too - clean parameter names, sensible defaults, clear required vs optional. It's basically UX design for machines rather than humans.

I don't follow. Wouldn't you do all those things to design for humans anyway?

Re: Ask HN: How do you know if AI agents will choose your tool?

#5

Post by LetsAutomate » Mon, Feb 23, 2026, 8:55 PM UTC

[flagged]

Re: Ask HN: How do you know if AI agents will choose your tool?

#6

Post by sincerely » Mon, Feb 23, 2026, 9:14 PM UTC

LetsAutomate wrote: post #5

[flagged]

You'd know, huh?

Re: Ask HN: How do you know if AI agents will choose your tool?

#7

Post by GenericDev » Mon, Feb 23, 2026, 9:41 PM UTC

JacobArthurs wrote: post #3

Tool description quality matters way more than people expect. In my experience with MCP servers, the biggest win is specificity about when not to use the tool. Agents pick confidently when there's a clear boundary, not a vague capability statement.

[dead]

Re: Ask HN: How do you know if AI agents will choose your tool?

#8

Post by snowhale » Mon, Feb 23, 2026, 10:08 PM UTC

[dead]

Re: Ask HN: How do you know if AI agents will choose your tool?

#9

Post by kellkell » Mon, Feb 23, 2026, 10:59 PM UTC

CRIPIX seems to be a new and unusual concept. I came across it recently and noticed it’s available on Amazon. The description mentions something called the Information Sovereign Anomaly and frames the work more like a technological and cognitive investigation than a traditional book. What caught my attention is that it appears to question current AI and computational assumptions rather than promote them. Has anyone here heard about it or looked into it ?

Re: Ask HN: How do you know if AI agents will choose your tool?

#10

Post by alexandroskyr » Tue, Feb 24, 2026, 1:03 AM UTC

Curious if anyone has seen differences in how models handle conflicting tool descriptions — e.g., two tools with overlapping capabilities where the boundary isn't clear. In my experience that's where most bad tool calls come from, not from missing descriptions but from ambiguous overlap between tools.