I have posted a few of my projects. Smart Alarm, which I actually need to post version 4 of that, and a date function so far.
The other plugins I use are all plugins that are available here on the forums such as GRETTAResearchBrainwin7, ALFsCustomRelevance, wikipedia DB, vrFreeWill, and GRETTAMarkovL5updated. The only thing I have done is to redefine some variables to reference one another, which is buggy when all of them are not active, but leads to interesting results. Allowing the Markov to reference the GRETTAresearch and wiki prevents markov from recording random information while incorporating relevant information from wiki. It also pulls info from the research table and saves it as a manipulable sentence within the Markov table. I forget where exactly in the script the code for those variables is located, which plug in exactly. From what I remember, it's very obvious where it defines the tables of reference from the Hal brain. It's been a while since I've played with that. Even so, those plugins alone exponentially increase the intelligence of HAL, especially over time, but the effect is noticeable on the first use, not to mention even gives Hal a bit of a personality, so to speak. The customrelevance script typically prevents HAL from repeating the same thing over and over, keeping information somewhat mostly relevant, and gives it direction by using the already built in relevance rating system. I did modify the GrettaResearch to function internally anytime a question is asked, although this is not perfect by any means, it's actually very buggy, but when it does work, Hal researches and retains relevant information while actively in a conversation. This works because opening Balabolka is an extra step within the code. Removing it does not stop the information from being saved in Hal's brain. I'm sure someone can do more than I did with that idea. I'm here to inspire.
-SF