Faraday Agents Dispatcher helps user develop integrations with Faraday written in any language.

Just run pip3 install faraday_agent_dispatcher and you should see the faraday-dispatcher command in your system.

To setup a development environment (this is, to change code of the dispatcher itself, not to write your own integrations), you should clone this repo and run pip install -e ..

1. Generate a configuration file running faraday-dispatcher config-wizard.

2. Run the agent with faraday-dispatcher run command. The config file that it creates will be located in ~/.faraday/config/dispatcher.ini if you do not pass a custom path.

You should complete the agent configuration with your registration token, located at http://localhost:5985/#/admin/agents. Check that the server section has the correct information about your Faraday Server instance. Then, complete the agent section with the desired name of your agent. Finally, add the executors

An executor is a script that prints out single-line JSON data to stdout. Remember that if you print any other data to stdout, the dispatcher will trigger an error. If you want to print debugging or logging information you should use stderr for that.

Every line written to stdout by the executor will be decoded by the dispatcher and sent to Faraday using the Bulk Create endpoint. Therefore, the JSON you print must have the schema that the endpoint requires (this schema is detailed below). Otherwise, the dispatcher will complain because you supplied invalid data to it.

If you want to debug your executor, the simplest way to do it is by running it directly instead of running with the Dispatcher. Since the executor just prints JSON data to stdout, you will be able to see all information it wants to send to Faraday, but without actually sending it.

After writing your executor, you have to add it with the faraday-dispatcher config-wizard within the executor section, adding its name, command to execute and the max size of the JSON to send to Faraday Server. Additionally, you can configure the Environment variables and Arguments in their proper section.

To run an executor use the faraday-dispatcher config-wizard command, and play it from the Faraday Server. The executor will use the environment variables set and ask for the arguments.

The data published to faraday must correspond to the bulk_create endpoint of the Faraday's REST API

If you want to have more than one dispatcher, each one runninng its own executors, the preferred of doing this is to create different configuration files for each one (for example, ~/.faraday/config/dispatcher-1.ini and ~/.faraday/config/dispatcher-2.ini). Then, you can run two different Dispatcher instances with faraday-dispatcher --config-file PATH_TO_A_CONFIG_FILE.

Inside the executors directory you can find the already created executors.

The official executors are the collection of ready-to-go executors (with minimum configuration with the wizard). They have a manifest JSON file, which gives details about the uses of the executor and helps with the configuration of them.

The current official executors are:

• Arachni
• Burp
• CrackMapExec
• Nessus
• Nikto
• Nmap
• Nuclei
• Openvas
• Report processor: Consumes a local report where the dispatcher is with the faraday plugins
• [QualysGuard] qualys
• Sonar Qube API
• Sublist3r
• W3af
• Wpscan
• Zap

The development executors are the collection of executors we do not fully maintain, we have examples of use, conceptual, and in development executors. The most important of them are:

• basic_example.py: The Hello World of Faraday executors. It will create a host with an associeted vulnerability to it
• heroku_discovery_agent.py: Load host and service information from your Heroku account
• prowlerSample.py: Run the prowler command and send its output to Faraday
• brainfuck.sh: A proof-of-concept to demonstrate you can create an executor in any programming language, including Brainfuck!

We are currently working on new executors, apart from improving the experience using the agents.

We would like to give some agents read access to their workspace, so they can benefit of the existing data in order to find more valuable information.

For more info you can check our documentation