Command Line Options:

-A
Return all results, no matter how many
-G
Return Query results in ARC format.
-H
Return Query results in a format suitable only for humans.
-L
Return Query results in LOCATION format.
-N
Do not do any relevance ranking. Often used with -A. Faster if you don't need it.
-P
Strip out unprintable characters when show file contents
-Q
Quiet. Do not emit the human readable kibitzing that otherwise appears on stderr.
-T
Return Query results as just titles.
-U
Return Query results in URL format.
-<MaxResults>
Return no more than this many results. Default 10.
<buffer-count>
how many 64K blocks should be cached in memory. A larger number slows down startup a little and might make things go faster when used in interactive mode.
Run Time Commands:

?
Help
,<expander>
A Query preceeded by a comma returns a list of words matched by the first term of the query. Whereas "stuff*" would return all pages with words beginning with "stuff", ",stuff" returns all the unique words (sorted alphabetically) beginning with "stuff".
.<command>
one of the commands discussed below
(query)
A query assumed to be in the Advanced Algebraic Query language
query
A query assumed to be in the Basic Query language
.alts [ []]
Display RecordID's for a particular Canon
.altdata [start [count]]
Display Alt Data for RecordIDs in the given range
.count <term>
Count the occurances of <term> in each field.
.dump
Show basic stats for the index
.files [start [count]]
Display information about FileIDs in the given range.
\t malt [start [count]]
Display Map information about Alt Index Blocks in the given range
.maltdata [start [count]]
Display Map information about Alt Data Blocks in the given range.
.mfiles [start [count]]
Display Map information about File Blocks in the given range.
.mposts [start [count]]
Display Map information about Posting Blocks in the given range.
.mpures [start [count]]
Display Map information about Pure Url Blocks in the given range.
.mrecords [start [count]]
Display Map information about record Blocks in the given range.
.mrecs [start [count]]
Display Map information about Small Record Blocks in the given range.
.parse <record id>
Display a word/field pair for each indexable term in the Record.
.posts <term> [<first record> [<count>]]
Display Postings for a Term.
.pures [start [count]]
Display information about pure urls in the given range.
.purify <url>
Display results of purifying the url
.records [start [count]]
Display information about recordIDs in the given range
.show <record id>
Display the text of a record, includeing arc and HTTP headers.
.tokenize <string>
Chop the string into tokens.
.
A lone period (of EOF) will exit av_explore.

Example :
echo "(stuff and junk)" | av_explore -nag index | other_stuff.pl