web

v0.1.0 · compiled with katari 0.1.0

Overview

from the package README

web — fetch a web page as a Katari tool

A single module, web: an HTTP GET exposed as a tool the model can call. Pure Katari over the prelude's http.fetch. No FFI sidecar, and no capability to provide — it reads public pages, so there is no key and nothing to use.

web.fetch_page(url) — the tool: fetches the page and returns its body text, truncated to roughly a page so several fetches in one turn do not blow the context window. A non-2xx status comes back as text (the model can react); only a request that never completes throws (http.fetch_error).

Secrets / env

None.

Usage

import web
 
agent read(url: string) -> string with io {
  web.fetch_page(url = url)
}

Hand web.fetch_page to an AI loop's tool list to let the model read pages on its own.

web

2 declarations

fetch_page#

agent

agent fetch_page(url: string) -> string

Tool: fetch a web page over HTTP GET and return its body text, truncated so the result stays small enough to feed back to the model. A non-2xx status is reported as text (the model can react); only a request that never completes throws (http.fetch_error).

Parameters

url

string

Returns

string

Inferred type

agent(url: string) -> string with throw[fetch_error] | io

Wire view (JSON Schema)

truncate#

agent

agent truncate(text: string) -> string

Parameters

text

string

Returns

string

Wire view (JSON Schema)