What you do need to know before you enter a graduate program in archaeology.

Previously, I argued that traditional field skills are not anything to worry about before entering a graduate school: you can always learn how to dig a hole. Now, you might ask, what skills should I have? Here's the list, based on what I've seen graduate students struggle with during their first years. Having these skills under control will greatly ease your first years and give you time to focus on learning actual archaeology rather than basic mechanics. You might also check out my previous post on my advice for graduate students.
(1) Be able to write. The better you are at writing, the easier your graduate school experience will be, especially in the first few years. In all of your classes you will be asked to write: essays, papers, laboratory write ups, analyses, reviews, etc. Writing is a key part of graduate school. In fact, its basically the product of archaeology. At our core, we are writers who happen to rummage around in dirt, speak conceptually about rocks and other kinds of inanimate things. The better you are at writing, the better an archaeologist you will be. Before you start graduate school, do some writing. Start a blog and keep it current. Turn a paper you wrote as an undergraduate into something for publication - whether that is a newsletter, local publication, regional venue or greater. Take a class in writing at your local community college. Read Strunk and White's Elements of Style. Then read it again. Then read On Writing Well, by William Zinsser. Then practice some more. Make your writing logically sound. Clear. Sharp. Readable.
(2) Learn how to use Microsoft Excel or any equivalent spreadsheet program (whether that be OpenOffice, Gnumeric, Numbers or some other variant). Given that contemporary students are often touted as being terrifically "computer savvy" with all of that YouFace, FaceTube and Twitface stuff, I am incredibly appalled that many entering archaeology students have never used a spreadsheet. I've seen students, for example, enter numbers into a spreadsheet as instructed and then do calculations on a calculator (e.g., summing the values) and then entering the results back into the spreadsheet. Really. Given that much of what we do in archaeology in addition to writing is counting things, have advanced knowledge of Excel will go along way to making your assignments be about the problems posed and not the mechanics of dealing with numbers. Trust me on this one. Learn how to use a spreadsheet, create headers, use autofilter, use functions, create formulas, create graphs, do basic statistics, etc. Equally important: learn how to edit the graphs and tables produced by Excel to make them acceptable. Don't rely on the defaults. Learn to identify "chartjunk" and maximize the information to ink ratios for your graphics.

(3) Know the basics of statistics. If you count, describe or measure things, you need to understand statistics simply because all of the actions are meaningful only if one understands the contribution of error. Don't hide behind some claim of "phobia" about numbers: if this is really the case, choose another career. Really. If you really are unable to understand the concept of a standard deviation, do society a favor and pick a more suitable career like crematorium manager or whatever. And while you will learn more about statistics in graduate school, if you don't know anything about the basics you will be wasting everyone else's time as you get up to speed. Take an extension class. Read a book on statistics and do the exercises. Practice. Count things, calculate error. Just learn statistics.

(4) Learn some kind of programming language. Some might argue with me about the fundamental nature of this skill, but I definitely believe that every archaeologist should learn at least on kind of programming language: Perl, Basic, Visual Basic, Python, C, C++, C#, shell scripting, Java, Javascript. It really does not matter which language one chooses, but the skills acquired through learning how to take a problem and break it down into simple steps that can be coded will provide a generous payoff for anyone tackling the routine problems facing archaeology students (and professionals alike). Programming is more of a logic problem than anything else and learning how to conceive of problems that can be tackled via a program is a good demonstration that you can figure out how to solve problems in general. And having the skills to program means you are not going to waste your life doing tedious things that you (a) not do well and (b) you shouldn't be doing by hand in the first place. For example, given a large number of files (say 10,000) rename them according to some scheme. If your only solution is to do this take one by one and by hand, you are in trouble. Take a class. Read a book. Do online exercises. Learn some kind of programming. And do it before you are in graduate school.
(5) Learn something about GIS. In graduate school, you will definitely have opportunities to learn something about spatial systems. However, you should gain a background in the basics of geographic systems before you start graduate school. GIS is really just a different kind of computer application - like Word or Excel - that allows you to manipulate concepts in space. No archaeologist worth her/his salt should lack an understanding of GIS. And that means you. The more you know about GIS - like using Word or Excel, the easier it will be to focus on archaeological analyses that deal with space rather than the mechanics of the various program that you are going to use to study space. Learn ESRI's ArcGIS or the open source Quantum GIS or one of many other geographic information systems.
Well, that is a start that should keep you busy for the summer prior to starting a graduate program in archaeology. As I recall other critical skills, I'll update this list. If you have any ideas, please let me know.