Editors supervise the set of sites monitored by Attribyte, adding new sites as they are discovered and occasionally removing those that drift off-topic or stop updating. To work with Attribyte, sites must publish a feed to which access is allowed by the robot exclusion standard (robots.txt) Each hour, all entries that have been published in the past seven days are indexed and scored by an algorithm. When this is complete, the sites are ranked based on the total score assigned to their entries. The top-ranked sites appear in the left-hand column, along with each identified author's contribution to the site's rank. The following diagram shows details of the site's components.

1. The site's title. Obtained either from the home page, or from the site's feed. May be edited for brevity. Links to a page of recent top-ranked entries for the site.
2. If multiple authors are identified, each author's name links to a page of the author's recent top-ranked entries.
3. The percent of the site's total rank contributed by entries attributed to the author.
4. The site's 'favicon' - periodically refreshed.

Entries link to blog posts or news articles. They consist of a headline, a snippet of text, an image thumbnail (when available) and a cluster of links that contribute to the entry's ranking. The following diagram describes the components of an entry as it appears on Attribyte.

1. The entry headline. Links to the full article as published on the original site.
2. A permanent link to the entry as it appears on Attribyte.
3. The author of the entry, if identified in the feed. Links to a page showing recent top-ranked entries for this author.
4. Identifies the source site. Links to the site's home page.
5. A snippet. Usually the first sentence of the article.
6. A thumbnail of the first image in the entry. Links to the full article containing the image. Images are processed for all ranked entries when they include images and their access is not restricted by robots.txt.
7. A "tree" of links related to the entry. The arrow identifies the link direction. Entries may have multiple links in and out, but Attribyte's algorithm builds a spanning tree, so links only appear once. Entries may appear without a tree if the links that contribute to their score are front page links, links from "round-up" entries or links from entries with robots.txt restrictions.
8. Links identified by Attribyte's software as the "first mentioned" or original source of information. The link leads to a site that may not be monitored by Attribyte.
9. A link to the external site. If an editor determines that the site is on-topic and there are no technical reasons preventing it from working with Attribyte (e.g., it has no feed), it may be added to Attribyte's set of actively monitored sites.
10. When a tree becomes too large, only a portion is shown. This link will expand the full tree.

Entries are assigned scores based on their inbound links. Front page links are considered as well as links found in entries. Each link contributes a different amount to the total score based on a number of factors including the following:

• The "page rank" of the sites participating in the link relationship - computed for the graph defined by links discovered on the front page for each site.
• The link history. If a site frequently links to another, the contribution to the total score is reduced. Conversely, a link from a site that has never, or rarely linked to another counts more.
• The total number of links from an entry. For example, some sites periodically publish "round-up" entries that contain many links. This reduces the value of each link.

The search index contains all entries, including those that are unranked. By default, the following constraints are applied to the terms in the search box:

• Matching is restricted to the title and the first few sentences.
• Only current entries are returned.
• Entries may contain any of the search terms.

• Exact phrases are matched by enclosing the phrase in quotes, e.g. "online storage"
• Multiple terms may be combined with the boolean operators: AND, OR and NOT, e.g. apple AND laptop
• Adding the prefix title: to any term matches only within titles.
• Adding + before a term requires it to appear in the result.
• Adding - before a term prohibits it from appearing in the result.

Popular topics are automatically identified by software that analyzes the titles and text of the current entries. Topics appear below the search box and link to pages that contain related entries, both ranked and unranked.

All entry pages, except search results, have a feed that mirrors the content on the page. To make search feeds less noisy, they contain only recent, ranked entries that match the search expression. The feed icon, may be used to subscribe to feeds. Feeds are also identified with tags for "auto-discovery."