rag-query(1)
==========

NAME
----
rag-query - Query on knowledge-bases

SYNOPSIS
--------
[verse]
'rag query' <query> [--json | -j] [--model <model>] [--schema <schema>]
            [--max-summaries <n>] [--max-retrieval <n>] [--enable-ii | --disable-ii]
            [--enable-rag | --disable-rag] [--super-rerank | --no-super-rerank]

'rag query' --interactive | -i | --multi-turn [--model <model>] [--max-summaries <n>]
            [--max-retrieval <n>] [--enable-ii | --disable-ii] [--enable-rag | --disable-rag]
            [--super-rerank | --no-super-rerank]

DESCRIPTION
-----------
Ask AI about the knowledge-base.

You can override the query configurations with cli options. For example,
if you run `rag query --max-retrieval=5 <QUERY>`, it will retrieve at most
5 chunks regardless of the configuration. It's temporary, it doesn't write
to the config file.

If `--json` is set, it dumps the result as a json. The json contains AI's
response and retrieved chunks. `--json` option and `--schema` option are
very different. In most cases, you don't need `--json` when `--schema` is
enabled.

`--schema` allows you to use pdl schemas with this command. Run
`rag help pdl-format` to learn more about pdl schemas. It forces LLM's output
schema. For example, `rag query <query> --schema="[int]"` will dump an array
of integers to stdout. You can pipe that array to another program.
If `--schema` is enabled and the schema is json, like `[int]` or
`{ name: str, age: int }`, you can always pipe stdout to a json processor.

If the LLM is not smart enough or request is not clear, it may fail to
generate a response with valid schema. In that case, it'll dump `null`
to stdout, which is still a valid json. You always have to check the output
because LLMs can always fail.