Skip to main content

IRC bot - full featured, yet extensible and customizable

Project description

-- restructuredtext --


pmxbot is an IRC bot written in python. Originally built for internal use, it’s been sanitized and set free upon the world. You can find out more details on the website,, and especially the wiki


pmxbot listens to commands prefixed by a ‘!’ If it’s a command it knows it will reply, take an action, etc. It can search the web, quote you, track karma, make decisions, and do just about anything else you could want. It logs text in a sqlite3 database, and eventually we’ll write a web interface to it.


pmxbot will respond to things you say if it detects words and phrases it’s been told to recognize. For example, mention sql on rails.


pmxbot requires python, of course. Probably 2.5+. It also requires a few python packages:

  • python-irclib
  • pyyaml
  • json (standard library for 2.6+) or simplejson
  • feedparser
  • httplib2
  • sqlite3

The web interface requires cherrypy and jinja2.

Optionally nosetests for the test suite.


We have a basic nose test suite that does some functional tests with an included TCL IRC daemon, and some basic unit tests as well. Just run them from the pmxbot root directory with “nosetests” and it should do it all for you. You’ll need TCL and nosetests to do it though.


Configuration is based on very easy YAML files. Check out config.yaml in the source tree for an example.


Once you’ve setup a config file, you just need to call pmxbot config.yaml and it will join and connect. We recommend running pmxbot under djb’s daemontools to make it automatically restart if it crashes.

Custom Features

Setuptools Entry Points Plugin

To create a setuptools (or distribute or compatible packaging tool) entry point plugin, package your modules using the setuptools tradition and install it alongside pmxbot. Your package should define an entry point in the group pmxbot_handlers by including something similar to the following in the package’s

entry_points = {
    'pmxbot_handlers': [
        'plugin_name = mylib.mymodule:initialize_func',

During startup (and after loading the traditional script-based plugins), pmxbot will load mylib.mymodule and call initialize_func from that module (with no parameters). plugin_name can be anything, but should be a name suitable to identify the plugin. It is not necessary that initialize_func do anything at all, but it must be present, and will be called during startup.

Within the script you’ll want to import the decorates you need to use with: from pmxbot.botbase import command, contains, execdelay, execat. You’ll then decorate each function with the appropriate line so pmxbot registers it.

A command (!g) gets the @command decorator:

@command("tinytear", aliases=('tt', 'tear', 'cry'), doc="I cry a tiny tear for you.")
def tinytear(client, event, channel, nick, rest):
      if rest:
              return "/me sheds a single tear for %s" % rest
              return "/me sits and cries as a single tear slowly trickles down its cheek"

A response (when someone says something) uses the @contains decorator:

def yay_sor(client, event, channel, nick, rest):
      karmaChange(botbase.logger.db, 'sql on rails', 1)
      return "Only 76,417 lines..."

For an example of how to implement a setuptools-based plugin, see the wolframalpha plugin.

Web Interface

pmxbot includes a web server for allowing users to view the logs, read the help, and check karma. You specify the host, port, base path, logo, title, etc with the same YAML config file. Just run like pmxbotweb config.yaml and it will start up. I suggest using a tool like supervisor, daemontools, cherryd or somesuch to keep it running.

Project details

Release history Release notifications

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Filename, size & hash SHA256 hash help File type Python version Upload date (2.1 MB) Copy SHA256 hash SHA256 Source None

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN SignalFx SignalFx Supporter DigiCert DigiCert EV certificate StatusPage StatusPage Status page