---
title: "Agent Quick Index"
author: "Rainer Krug"
format: html
vignette: >
  %\VignetteIndexEntry{Agent Quick Index}
  %\VignetteEngine{quarto::html}
  %\VignetteEncoding{UTF-8}
execute:
  echo: true
  warning: false
  message: false
  eval: false
---

This page is a machine-friendly, human-readable quick selector for exported
functions.

## Connection and Authentication

- `kagi_connection()`
  Build a reusable authenticated connection.

## Query Construction

- `query_search()`
  Build Search endpoint requests.
- `query_enrich_web()`
  Build Enrich Web endpoint requests.
- `query_enrich_news()`
  Build Enrich News endpoint requests.
- `query_summarize()`
  Build Summarize endpoint requests.
- `query_fastgpt()`
  Build FastGPT endpoint requests.

## Request Execution

- `kagi_request()`
  Low-level request execution into JSON output folders.
- `kagi_fetch()`
  High-level project workflow executor writing endpoint-scoped JSON/parquet.

## Storage and Refresh

- `kagi_request_parquet()`
  Convert request JSON folders to parquet.
- `kagi_update_query()`
  Re-run one stored query by name and refresh touched parquet partitions.
- `clean_request()`
  Remove JSON payload artifacts while preserving query metadata.

## Content and Abstract Pipeline

- `download_content()`
  Download source content for endpoint/query partitions.
- `content_markdown()`
  Convert downloaded files to markdown.
- `markdown_abstract()`
  Summarize markdown and write abstract parquet outputs.
- `summarize_with_openai()`
  OpenAI text summarization provider.
- `summarize_with_kagi()`
  Kagi text summarization provider.
- `read_corpus()`
  Read parquet corpus and optionally link abstract data.

## Utility

- `open_search_query()`
  Open rendered search query in browser for inspection.