Skip to content

Read indexed content

Search returns matching documents and snippets. Use the read commands when you want the indexed document text. For extracted formats such as HTML and digital PDFs, this is the canonical text stored in the index rather than raw file bytes.

List files in a collection

kbolt ls my_docs
kbolt ls my_docs guides/

Use a prefix to narrow the listing to one part of the collection.

Read one file

Use a collection-relative path:

kbolt get my_docs/path/to/file.md

Or use a document id from search output:

kbolt get '#abc123'

Read a line range with --offset and --limit:

kbolt get my_docs/path/to/file.md --offset 40 --limit 80

Read several files

Use multi-get when an agent or script needs several documents in one call:

kbolt multi-get my_docs/a.md,my_docs/b.md

Set limits when you need bounded output:

kbolt multi-get my_docs/a.md,my_docs/b.md --max-files 10 --max-bytes 30000

When to use each command

Need Command
See what is indexed kbolt ls
Read one document kbolt get
Read several documents with output limits kbolt multi-get
Let an agent call read tools directly kbolt mcp

Next steps