- "Replies can be prefixed by 'act:' (e.g. act:goes shopping) to signify that the bot should act instead of saying the message. " +
- "Replies can use the %%{key} syntax to access one of the following keys: " +
- "who (the user that said the trigger), bot (the bot's own nick), " +
- "target (the first word following the trigger), what (whatever follows target), " +
- "stuff (everything that follows the trigger), match (the actual matched text)"
+ "Replies prefixed by 'act:' (e.g. act:goes shopping) signify that the bot should act instead of saying the message. " +
+ "Replies prefixed by 'cmd:' or 'command:' (e.g. cmd:lart %{who}) issue a command to the bot. " +
+ "Replies can use the %{key} syntax to access the following keys: " +
+ "who (user that said the trigger), bot (bot's own nick), " +
+ "target (first word following the trigger), what (whatever follows target), " +
+ "before (everything that precedes the trigger), after, (everything that follows the trigger), " +
+ "match (matched text), match1, match2, ... (the i-th capture). " +
+ "Replies prefixed by 'ruby:' (e.g. ruby:m.reply 'Hello ' + subs[:who]) are interpreted as ruby code. " +
+ "No %{key} substitution is done in this case, use the subs hash in the code instead. " +
+ "Be warned that creating ruby replies can open unexpected security holes in the bot."