Welcome to the VirusTotal CLI, a tool designed for those who love both VirusTotal and command-line interfaces. With this tool you can do everything you'd normally do using VirusTotal's web page, including:
- Retrieve information about a file, URL, domain name, IP address, etc.
- Search for files and URLs using VirusTotal Intelligence query syntax.
- Download files.
- Manage your LiveHunt YARA rules.
- Launch Retrohunt jobs and get their results.
And much more...
As this tool uses the VirusTotal API under the hood, you will need a VirusTotal API key. By signing up with VirusTotal you will receive a free API key however, free API keys have a limited amount of requests per minute, and they don't have access to some premium features like searches and file downloads. If you are interested in using those premium features please contact us.
There are two ways of installing the tool: by using one of our pre-compiled binaries or by building it by yourself.
The pre-compiled binaries can be found on the releases page. There are binaries for Windows, Linux and Mac OS X. To use them, just download the file, decompress it and place it in a directory where you think is more convenient to use.
To compile the program you'll need Go 1.14.x or higher installed in your system and type the following commands:
$ git clone https://github.com/VirusTotal/vt-cli
$ cd vt-cli
$ make install
NOTE: in order to use the vt
binary, make sure the GOBIN
is part of your PATH
env variable:
$ export GOBIN=`go env GOPATH`/bin
$ export PATH=$PATH:$GOBIN
For Mac OS users, there's a brew formula available. Please note this is not maintained by VirusTotal.
$ brew install virustotal-cli
For Windows users, there's a Winget manifest available. Please note this is not maintained by VirusTotal.
winget install VirusTotal.vt-cli
Chocolatey is also supported (Also not maintained by VirusTotal):
choco install vt-cli
If you plan to use vt-cli
in Windows on a regular basis we highly recommend you avoid the standard Windows console and use Cygwin instead. The Windows console is very slow when printing large amounts of text (as vt-cli
usually does) while Cygwin performs much better. Additionally, you can benefit from Cygwin's support for command auto-completion, a handy feature that the Windows console doesn't offer. In order to take advantage of auto-completion make sure to include the bash-completion
package while installing Cygwin.
Once you have installed the vt-cli tool you may want to configure it with your API key. This is not strictly necessary, as you can provide your API key every time you invoke the tool by using the --apikey
option (-k
in short form), but that's a bit of a hassle if you are going to use the tool frequently (and we bet you'll do!). For configuring your API key just type:
$ vt init
This command will ask for your API key, and save it to a config file in your home directory (~/.vt.toml). You can also specify your API key using the VTCLI_APIKEY
environment variable. If you specify your API key in multiple ways, the --apikey
option will have the highest precedence, followed by the VTCLI_APIKEY
environment variable, the API key in the configuration file will be used as the last resort.
If you are behind an HTTP proxy you can tell vt-cli
which is the address of your proxy server in multiple ways. One is using the --proxy
option, like in:
$ vt --proxy http://myproxy.com:1234 <command>
You can also use the VTCLI_PROXY
environment variable, or add the following line to the config file:
proxy="http://myproxy.com:1234"
If you are going to use this tool frequently you may want to have command auto-completion. It saves both precious time and keystrokes. Notice however that you must configure your API as described in the previous section before following the steps listed below. The API is necessary for determining the commands that you will have access to.
-
Linux:
$ vt completion bash > /etc/bash_completion.d/vt
-
Mac OS X:
$ brew install bash-completion $ vt completion bash > $(brew --prefix)/etc/bash_completion.d/vt
Add the following lines to
~/.bash_profile
:if [ -f $(brew --prefix)/etc/bash_completion ]; then . $(brew --prefix)/etc/bash_completion fi
-
Cygwin:
Make sure the
bash-completion
package is installed (Cygwin doesn't install it by default) and type:$ vt completion bash > /usr/share/bash-completion/completions/vt
❗ You may need to restart your shell in order for autocompletion to start working.
The output script from vt completion zsh
needs to be put somewhere under the $fpath
directory. For example, .oh-my-zsh/completions
directory:
$ mkdir /Users/$USERNAME/.oh-my-zsh/completions
$ vt completion zsh > /Users/$USERNAME/.oh-my-zsh/completions/_vt
Restart the shell.
-
Get information about a file:
$ vt file 8739c76e681f900923b900c9df0ef75cf421d39cabb54650c4b9ad19b6a76d85
-
Get information about a file in JSON format:
$ vt file 8739c76e681f900923b900c9df0ef75cf421d39cabb54650c4b9ad19b6a76d85 --format json
-
Get a specific analysis report for a file:
$ # File analysis IDs can be given as `f-<file_SHA256_hash>-<UNIX timestamp>`... $ vt analysis f-8739c76e681f900923b900c9df0ef75cf421d39cabb54650c4b9ad19b6a76d85-1546309359 $ # ...or as a Base64 encoded string, retrieved from the `vt scan file` command: $ vt scan file test.txt test.txt MDJiY2FiZmZmZmQxNmZlMGZjMjUwZjA4Y2FkOTVlMGM6MTU0NjQ1NDUyMA== $ vt analysis MDJiY2FiZmZmZmQxNmZlMGZjMjUwZjA4Y2FkOTVlMGM6MTU0NjQ1NDUyMA== - _id: "MDJiY2FiZmZmZmQxNmZlMGZjMjUwZjA4Y2FkOTVlMGM6MTU0NjQ1NDUyMA==" _type: "analysis" date: 1546454520 # 2019-01-02 13:42:00 -0500 EST stats: failure: 0 harmless: 0 malicious: 0 suspicious: 0 timeout: 0 type-unsupported: 0 undetected: 0 status: "queued"
-
Download files given a list of hashes in a text file, one hash per line:
$ cat /path/list_of_hashes.txt | vt download -
-
Get information about a URL:
$ vt url http://www.virustotal.com
-
Get the IP address that served a URL:
$ vt url last_serving_ip_address http://www.virustotal.com
-
Search for files:
$ vt search "positives:5+ type:pdf"
-
Scan a file:
$ vt scan file <yourfile> <yourfile> ZDZiOTcxY2JhNDE0MWU5ZWRjN2JjNGQ2NTdhN2VjODU6MTU3MDE3Mjg1NQ== $ vt analysis ZDZiOTcxY2JhNDE0MWU5ZWRjN2JjNGQ2NTdhN2VjODU6MTU3MDE3Mjg1NQ== - _id: "ZDZiOTcxY2JhNDE0MWU5ZWRjN2JjNGQ2NTdhN2VjODU6MTU3MDE3Mjg1NQ==" _type: "analysis" date: 1570172855 # 2019-10-04 09:07:35 +0200 CEST stats: failure: 0 harmless: 0 malicious: 0 suspicious: 0 timeout: 0 type-unsupported: 0 undetected: 0 status: "queued"
-
Export detections and tags of files from a search in CSV format:
$ vt search "positives:5+ type:pdf" -i sha256,last_analysis_stats.malicious,tags --format csv
-
Export detections and tags of files from a search in JSON format:
$ vt search "positives:5+ type:pdf" -i sha256,last_analysis_stats.malicious,tags --format json
When you ask for information about a file, URL, domain, IP address or any other object in VirusTotal, you get a lot of data (by default in YAML format) that is usually more than what you need. You can narrow down the information shown by the vt-cli tool by using the --include
and --exclude
command-line options (-i
and -x
in short form).
These options accept patterns that are matched against the fields composing the data, and allow you to include only a subset of them, or exclude any field that is not interesting for you. Let's see how it works using the data we have about http://www.virustotal.com
as an example:
$ vt url http://www.virustotal.com
- _id: 1db0ad7dbcec0676710ea0eaacd35d5e471d3e11944d53bcbd31f0cbd11bce31
_type: "url"
first_submission_date: 1275391445 # 2010-06-01 13:24:05 +0200 CEST
last_analysis_date: 1532442650 # 2018-07-24 16:30:50 +0200 CEST
last_analysis_results:
ADMINUSLabs:
category: "harmless"
engine_name: "ADMINUSLabs"
result: "clean"
AegisLab WebGuard:
category: "harmless"
engine_name: "AegisLab WebGuard"
result: "clean"
AlienVault:
category: "harmless"
engine_name: "AlienVault"
result: "clean"
last_http_response_code: 200
last_http_response_content_length: 7216
last_http_response_content_sha256: "7ed66734d9fb8c5a922fffd039c1cd5d85f8c2bb39d14803983528437852ba94"
last_http_response_headers:
age: "26"
cache-control: "public, max-age=60"
content-length: "7216"
content-type: "text/html"
date: "Tue, 24 Jul 2018 14:30:24 GMT"
etag: "\"bGPKJQ\""
expires: "Tue, 24 Jul 2018 14:31:24 GMT"
server: "Google Frontend"
x-cloud-trace-context: "131ac6cb5e2cdb7970d54ee42fd5ce4a"
x-frame-options: "DENY"
last_submission_date: 1532442650 # 2018-07-24 16:30:50 +0200 CEST
private: false
reputation: 1484
times_submitted: 213227
total_votes:
harmless: 660
malicious: 197
Notice that the returned data usually follows a hierarchical structure, with some top-level fields that may contain subfields which in turn can contain their own subfields. In the example above last_http_response_headers
has subfields age
, cache-control
, content-length
and so on, while total_votes
has harmless
and malicious
. For referring to a particular field within the hierarchy we can use a path, similar to how we identify a file in our computers, but in this case, we are going to use a dot character (.) as the separator for path components, instead of the slashes (or backslashes) used by most file systems. The following ones are valid paths for our example structure:
last_http_response_headers.age
total_votes.harmless
last_analysis_results.ADMINUSLabs.category
last_analysis_results.ADMINUSLabs.engine_name
The filters accepted by both --include
and --exclude
are paths in which we can use *
and **
as placeholders for one and many path elements respectively. For example foo.*
matches foo.bar
but not foo.bar.baz
, while foo.**
matches foo.bar
, foo.bar.baz
and foo.bar.baz.qux
. In the other hand, foo.*.qux
matches foo.bar.qux
and foo.baz.qux
but not foo.bar.baz.qux
, while foo.**.qux
matches
foo.bar.baz.qux
and any other path starting with foo
and ending with qux
.
For cherry-picking only the fields you want, you should use --include
followed by a path pattern as explained above. You can also include more than one pattern either by using the --include
argument multiple times, or by using it with a comma-separated list of patterns. The following two options are equivalent:
$ vt url http://www.virustotal.com --include=reputation --include=total_votes.*
$ vt url http://www.virustotal.com --include=reputation,total_votes.*
Here you have different examples with their outputs (assuming that vt url http://www.virustotal.com
returns the structure shown above):
$ vt url http://www.virustotal.com --include=last_http_response_headers.server
- last_http_response_headers:
server: "Google Frontend"
$ vt url http://www.virustotal.com --include=last_http_response_headers.*
- last_http_response_headers:
age: "26"
cache-control: "public, max-age=60"
content-length: "7216"
content-type: "text/html"
date: "Tue, 24 Jul 2018 14:30:24 GMT"
etag: "\"bGPKJQ\""
expires: "Tue, 24 Jul 2018 14:31:24 GMT"
server: "Google Frontend"
x-cloud-trace-context: "131ac6cb5e2cdb7970d54ee42fd5ce4a"
x-frame-options: "DENY"
$ vt url http://www.virustotal.com --include=last_analysis_results.**
- last_analysis_results:
ADMINUSLabs:
category: "harmless"
engine_name: "ADMINUSLabs"
result: "clean"
AegisLab WebGuard:
category: "harmless"
engine_name: "AegisLab WebGuard"
result: "clean"
AlienVault:
category: "harmless"
engine_name: "AlienVault"
result: "clean"
$ vt url http://www.virustotal.com --include=last_analysis_results.*.result
- last_analysis_results:
ADMINUSLabs:
result: "clean"
AegisLab WebGuard:
result: "clean"
AlienVault:
result: "clean"
$ vt url http://www.virustotal.com --include=**.result
- last_analysis_results:
ADMINUSLabs:
result: "clean"
AegisLab WebGuard:
result: "clean"
AlienVault:
result: "clean"
Also notice that _id
and _type
are also field names and therefore you can use them in your filters:
$ vt url http://www.virustotal.com --include=_id,_type,**.result
- _id: "1db0ad7dbcec0676710ea0eaacd35d5e471d3e11944d53bcbd31f0cbd11bce31"
_type: "file"
last_analysis_results:
ADMINUSLabs:
result: "clean"
AegisLab WebGuard:
result: "clean"
AlienVault:
result: "clean"
The --exclude
option works similarly to --include
but instead of including the matching fields in the output, it includes everything except the matching fields. You can use this option when you want to keep most of the fields, but leave out a few of them that are not interesting. If you use --include
and --exclude
simultaneously --include
enters in action first, including only the fields that match the --include
patterns, while --exclude
comes in after that, removing any remaining field that matches the --exclude
patterns.