This is a map of the codebase, sort of: http://mongobot.com/static/img/mongomap2.pdf

There's kind of a bootstrap.sh file that might work if you're on a Mac. No guarantees.

mkvirtualenv MongoBot
pip install -r requirements.txt

To run bot: python medulla.py

To run bot persistently so it will recover from crashes: python doctor.py

You'll need to install MongoDB to use most of the features relating to people. I apologize for this.

The master development branch is the skynet branch. There is a reason.

For best results in answering this question, read through medulla.py and cortex.py. The comments are... lengthy. The brainmeat interface is mostly in autonomic.py. Thalamus is what gets it on IRC and handles input. Also, learn python.

MongoBot is capable of running a number of web services and apis. For best results, run an nginx server and uwsgi server/deploy.ini. And by best results, I mean any. More documentation on this once I figure out how I did it the first time.

Also, the markov functionality is based on redis, so you'll have to get a redis server up and running.

You will need to create a file in the config folder called secrets.yaml. sample.secrets.yaml has descriptions of each setting, just copy it to secrets.yaml and set away.

To create a new command in an existing brainmeat category, add the function to the class and add the decorator @axon. To add a help entry for the function, add @help("Help message.")

To create a new command category, run this from the root directory:

python newbrains.py category_name

This will create a file called category_name.py in brainmeats, with a class Category_name. Initialization of this class will be automated once category_name is added to config/plugins.yaml.

If you actually manage to get this guy up and running, the default command characters are '.' for a command, and ':' for running multiple inputs to a command, so

.q AAPL

... gets you a stock quote for Apple, and

:q AAPL NFLX TWTR

... gets you quotes for Apple, Netflix, and Twitter.

When using in the chat room, you can pipe commands, i.e.:

.babble Zaphod | .tweet

... will tweet whatever babble spits out.

Why would anyone add piping to a chat bot? I dunno. Things got out of hand.

IMPORTANT: whoever is running the bot instance will have to use .adduser to add other users. After that they can authorize themselves.

Long, rambling comments are encouraged.

The brain puns SHALL be adhered to, to the very limits of linguistic feasibility.

Don't nest logic when you can short circuit.

BAD:

if blah:
    for x in stuff:
        do stuff

GOOD:

if not blah: return

for x in stuff:
    do stuff

Try to stick to single quotes wherever possible.

String interpolation is better than adding up strings. I'm totally not great at remembering this, but I'm trying.

If you want a function to be pipeable to other functions, return the output instead of just chatting from the function.

All commit messages for commits touching zalgo code must be 'ZALGO' or 'HE COMES', otherwise the pull request will be refused.

• Probably make this readme better
• turing.py function names need to be better, this is fucking mandarin
• some odd bugs in channeling
• remove image after thumbnailing, clear all out on boot
• remove video downloads on reboot
• create bootstrap install
• Add autostart for server, if components in place
• add tweet at in twitting
• reddit command breaks without specified subreddit
• Clean up broca
• link holdem to db, make persistant, open up sit/in/out functionality
• finish holdem, needs testing and split pots probably don't work
• also, holdem is just totally broken right now
• stock game: account for splits and reverse splits
• add @requires decorator, check for server, redis, mongo, etc.
• just take out image relink, make it a thumblink
• mud-style game, playable and modifiable from chat
• interbot com
• streamline web stuff
• multi-platform
• futher abstract thalamus to load an interface depending on connection type