There is a space at the bottom of the world, where Earth and Ice and Shadow meet. It is cold, but not cold enough to kill; dark, but not too dark to see. A small round room, made of chilly black marble, lit by a dim and sourceless glow, with a spiral stair climbing the curve of the wall and a shallow circular recession in the exact center of the floor. The recession is maybe six feet wide by six inches deep, lined with something resembling pale frosted glass, and there is nothing in it.
"True," he concedes. "I guess we'll see."
He opens the book and puts a few more remembered objects in it. "—Would you like a book to read? I have a much better selection of those than meals."
"I put in every book I remembered and I have lived in the palace all my life and I read a lot - I haven't actually tried but I'm pretty sure I got enough of the palace library's selection for the rest to fill itself in when I put the palace together, I could go in and pull one out for you."
"Somehow I suspected you'd go for that one."
He appears a book.
"There you go, have fun."
It's an introductory magical theory book, written in a cheerful, engaging style.
The basic principles of the system it describes are as follows:
A short, simple ritual involving no nonmagical components can imbue temporary magical effects into an object. The simplest effects are things like light and heat and cold and repetitive chiming sounds. Somewhat more complicated rituals get you temporary magical tools, which can be used in further rituals to give things permanent magical effects, including yet more tools, which you can use to give things fancier and more useful permanent effects, and so on, and so forth. To the best of anyone's knowledge there is no upper limit on the number of layers, except for practical ones like 'an eighty-year manufacturing time is probably too long to be worth bothering with'.
Rituals always result in persistent-if-not-permanent magical properties rather than one-time effects: there are no rituals that will turn uncarved wood into carved wood, or stale food into fresh, or anything like that.
You cannot make a living thing into a magical artifact; it just doesn't work. There's some speculation about why, and people have done experiments to narrow down exactly where the line falls between living and dead, but it's a pretty firm boundary even though there are a handful of edge cases (mostly involving plants).
Artifacts primarily affect themselves rather than the world around them. You can make a light-stone but you can't make a stone that causes objects in its vicinity to glow. You can make a self-turning water pump but you can't make a wheel that levitates water out of wells. The major exception is healing artifacts, which are a thriving field of research as people try to develop better ones and see if they can extend the concept into other domains. So far there's been some success in using healing-artifact techniques to develop artifacts that cause plants to flourish.
The entire second half of the book is devoted to the principles of ritual development. It's possible for multiple rituals to have the same result, but not for the same ritual to have multiple different results, so the first thing most people do when getting into ritual development is familiarize themselves with all the existing rituals in their area of study. Ritual development is a process of creation, not discovery: unless you happen to duplicate a previous invention, what you do when you develop a ritual is make that sequence of actions magical when they weren't before.
There's a lot of flexibility in the possible correspondences between ritual actions/components and the eventual result, but it's not completely freeform: the other advantage of getting familiar with the existing rituals in your chosen area is that you're almost always better off tweaking or combining those than trying to come up with something completely new. People with more experience in ritual development tend to be more successful at novel inventions, which many of them ascribe to simply having picked up a better feel for how ritual-building works, but there is a theory - unconfirmed but circumstantially supported - that successful ritual development involves a kind of unseen magical strength that increases as you do more of it.
Manipulation of nonliving objects, more or less. Anything that could be a valid ritual target is a valid ritual component and vice versa. Flexibility of timing varies according to a number of related factors including overall ritual complexity, how many different things need to be done in close succession at that stage of the ritual, and how 'messy' the steps are - for example, a step requiring you to rotate a marked disc counterclockwise might call for very precise timing, but a step requiring you to burn a folded paper will be more flexible because the time the paper takes to burn isn't going to be identical every time.
(Part of the skill of ritual invention is in specifying which objects and actions are and are not part of the ritual: thanks to a doubly careless inventor whose name has been lost to history, there exists a ritual for enchanting permanent heat-stones that requires you to be wearing a straw hat for most of the ritual, set the brim on fire while lighting the last round of candles, and then stamp it out and knock over half the candles in the process. It makes really good heat-stones but people don't use it very often, in part because it's hard to perform the frantic hat-stomping step to within tolerances. Also because it's silly. And a fire hazard.)
—he bursts out laughing. "I'd actually forgotten about that one, it's been years since I read the book."
"What a thing to be known for. I can't decide whether it's better or worse that we've forgotten his name."
"It caught fire. And was immortalized with its brethren as an ingredient in a heatstone ritual, catching fire and being stomped on and all."
"And that's why it's important to clearly specify your ritual actions and components!"
"We have some things a little like that but it's not challenging - if I paint blood all over a workspace and then leave the bucket, I am not going to accidentally incorporate the aesthetic of a bucket -"
"Well now I'm curious if anyone's ever managed to accidentally incorporate their bucket."