Output formats

Every command renders through one output layer, so the same data is a table for reading or a stream for piping. Choose with -o/--output.

The formats

-o value	What you get
table	aligned, human-readable columns (the default on a terminal)
json	one JSON array of objects
jsonl	one JSON object per line (newline-delimited)
csv	comma-separated, with a header row
tsv	tab-separated, with a header row
url	one URL per row (details, download, or replay, by command)
raw	the bare primary value per row (e.g. an identifier or file bytes)

auto

The default, auto, picks table when stdout is a terminal and jsonl when it is a pipe or file, so interactive use is readable and scripted use is machine-friendly without a flag:

archive search 'collection:nasa' -n 3            # table on screen
archive search 'collection:nasa' -n 3 | cat       # jsonl into a pipe

Projecting columns

--fields restricts and orders the columns by name, across every format:

archive search 'collection:nasa' --fields identifier,downloads
archive search 'collection:nasa' --fields identifier,downloads -o csv
archive files nasa --fields name,size -o tsv

--no-header drops the header row from table, csv, and tsv output, which is handy when feeding a column straight into another tool.

URLs and raw values for pipelines

-o url gives the natural URL for each row, and -o raw gives the bare primary value, both newline-delimited and perfect for xargs:

archive search 'collection:nasa' -o url | head
archive search 'collection:nasa' --fields identifier -o raw -n 5 | xargs -n1 archive item
archive files nasa --glob '*.jpg' -o url | xargs -n1 curl -sO

For wayback get, -o raw (the default) streams the snapshot bytes straight to stdout, so you can redirect them to a file or pipe them onward.

Templates

For full control, --template applies a Go text/template to each row's value object and prints the result:

archive search 'collection:nasa' -n 5 \
  --template '{{.identifier}} has {{.downloads}} downloads'

The template sees the same fields you would get from -o json, so inspect a row with -o json first to learn the field names, then template over them.