A keyword cipher is a form of monoalphabetic substitution. A keyword is used as the key, and it determines the letter matchings of the cipher alphabet to the plain alphabet. Repeats of letters in the word are removed, then the cipher alphabet is generated with the keyword matching to A,B,C etc. until the keyword is used up, whereupon the rest of the ciphertext letters are used in alphabetical order, excluding those already used in the key.
Plaintext A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
Encrypted K R Y P T O S A B C D E F G H I J L M N Q U V W X Z
So with KRYPTOS as the keyword, all A's become K's, all B's become R's and so on. Encrypting the message "cryptography is cool" using the keyword "kryptos":
Plaintext C R Y P T O G R A P H Y I S C O O L
Encoded Y L X I N H S L K I A X B M Y H H E
Only one alphabet is used here, so the cipher is monoalphabetic.
The best way to attack a keyword cipher is through a known-text attack, frequency analysis and discovery of the keyword (often a cryptanalist will combine all three techniques). Keyword discovery allows immediate decryption since the table can be made immediately. If you do not know the keyword, one of the main weaknesses of a keyword substitution is that it is vulnerable to attack by frequency analysis.