Multiple unspecified vulnerabilities in WordPress before 2.0.4 have unknown impact and remote attack vectors. NOTE: due to lack of details, it is not clear how these issues are different from CVE-2006-3389 and CVE-2006-3390, although it is likely that 2.0.4 addresses an unspecified issue related to "Anyone can register" functionality (user registration for guests).
index.php in WordPress 2.0.3 allows remote attackers to obtain sensitive information, such as SQL table prefixes, via an invalid paged parameter, which displays the information in an SQL error message. NOTE: this issue has been disputed by a third party who states that the issue does not leak any target-specific information.
WordPress 2.0.3 allows remote attackers to obtain the installation path via a direct request to various files, such as those in the (1) wp-admin, (2) wp-content, and (3) wp-includes directories, possibly due to uninitialized variables.
vars.php in WordPress 2.0.2, possibly when running on Mac OS X, allows remote attackers to spoof their IP address via a PC_REMOTE_ADDR HTTP header, which vars.php uses to redefine $_SERVER['REMOTE_ADDR'].
Direct static code injection vulnerability in WordPress 2.0.2 and earlier allows remote attackers to execute arbitrary commands by inserting a carriage return and PHP code when updating a profile, which is appended after a special comment sequence into files in (1) wp-content/cache/userlogins/ (2) wp-content/cache/users/ which are later included by cache.php, as demonstrated using the displayname argument.
Cross-site scripting (XSS) vulnerability in the paging links functionality in template-functions-links.php in Wordpress 1.5.2, and possibly other versions before 2.0.1, allows remote attackers to inject arbitrary web script or HTML to Internet Explorer users via the request URI ($_SERVER['REQUEST_URI']).
Multiple "unannounced" cross-site scripting (XSS) vulnerabilities in WordPress before 2.0.2 allow remote attackers to inject arbitrary web script or HTML via unknown attack vectors.
SQL injection vulnerability in WordPress 1.5.2, and possibly other versions before 2.0, allows remote attackers to execute arbitrary SQL commands via the User-Agent field in an HTTP header for a comment.
Multiple cross-site scripting (XSS) vulnerabilities in the "post comment" functionality of WordPress 2.0.1 and earlier allow remote attackers to inject arbitrary web script or HTML via the (1) name, (2) website, and (3) comment parameters.
WordPress 2.0.1 and earlier allows remote attackers to obtain sensitive information via a direct request to (1) default-filters.php, (2) template-loader.php, (3) rss-functions.php, (4) locale.php, (5) wp-db.php, and (6) kses.php in the wp-includes/ directory; and (7) edit-form-advanced.php, (8) admin-functions.php, (9) edit-link-form.php, (10) edit-page-form.php, (11) admin-footer.php, and (12) menu.php in the wp-admin directory; and possibly (13) list directory contents of the wp-includes directory. NOTE: the vars.php, edit-form.php, wp-settings.php, and edit-form-comment.php vectors are already covered by CVE-2005-4463. The menu-header.php vector is already covered by CVE-2005-2110. Other vectors might be covered by CVE-2005-1688. NOTE: if the typical installation of WordPress does not list any site-specific files to wp-includes, then vector [13] is not an exposure.