Between 1903 and 1907, G. Legrain discovered around 800 stone statues, stelae and other objects in a large pit (the so-called “Cachette”) inside the temple of Amun at Karnak, in which they were piously buried by the Egyptian priests, probably during the 1st century B.C. They include a number of royal effigies of all periods but most of the statues primarily belong to the priests who officiated at Karnak from the New Kingdom to the end of the Ptolemaic Period. The Karnak Cachette Database is an on-line inventory of the Cachette and a tool to search this rich corpus. The first version was launched in 2009; it provides, insofar as possible, a general description of each object (with dimensions, materials, dating), a label, the date of discovery, different inventory numbers, and a bibliography. Version 2 was put online in 2012: it includes an extensive access to the photographic documentation (more than 8,000 photographs are now available); this database has been regularly updated thereafter. Building on this well-defined corpus, the project aims now at developing the tools to encode, search and publish electronically the hieroglyphic texts inscribed on these objects, which provide anthroponomical, toponymical and prosopographical data and are therefore of historical and documentary significance. The encoding is developed according to the recommendations of the Text Encoding Initiative in combination with relevant “best practices” in the field of Digital Humanities applied to Epigraphy (Elliott et alii 2007; Cayless et alii, 2009). In this sense, even though the project takes into account many of the EpiDoc schema rules, it is only partially compliant with this TEI customization because of the specificities of both the project and Ancient Egyptian Epigraphy (compare with Lamé 2009), and also because there is a necessity to fall within the scope of other Egyptological projects dealing with textual corpora (Winand, Polis, Rosmorduc in press). Xefee, a tool to encode transliterated hieroglyphic inscriptions It is well known that XML is far from being a human friendly way to encode texts. Several XML editors are already available; some of them are highly customizable and can be used by very specific project, providing the users are proprely trained and some implementation time and effort is spent. However, due to the specific features of the texts from the Karnak Cachette – for instance in terms of prosopography –, and the general philosophy of the project – edit and analyse texts that require full Egyptological proficiencies –, it has been decided to create a specific XML Editor that would make easier the text input, its marking up as well as the generation of the XML/TEI files. Xefee – XML Editor for Egyptian Epigraphy – is a desktop Java application developed on Netbeans. It mainly consists of a general user interface (GUI) which provides all the necessary tools for managing and encoding the ancient Egyptian texts as well as the descriptive data pertaining to the Karnak Cachette project. These tools range from an import module that directly converts to XML the hieroglyphic text transcriptions written according to Egyptological standards, to more complex components intended to manage genealogical data. The tab dealing with text encoding offers to the user a panel of buttons, combo-boxes and other controls that facilitate the marking up the texts with tags pertaining to epigraphy (, , , , elements), onomastic ( element and elements with specific @type such as “deity”, “deityEpithet”, “toponym”) and prosopography ( elements with specific @type such as “person”, “title”, “filiationMark”). To add a tag, the user simply has to select in the top view pane the text to be marked up, and to press the appropriate button on the right-hand half of the tab. Since the XML marking up can be quite dense, mainly because the texts the project is dealing with often consist in compact sequences of personal names and titles, a preview pane in the bottom of the tab renders the encoded strings with different kinds of surrounding or highlighting patterns. The Ancient Egyptian way to present genealogical filiations also required to build up peculiar tools to handle this very important aspect of the text contents. A tab of the GUI is dedicated to the creation of person’s identities, whilst another one intends to manage the family links and generate the element. The current stage of the Karnak Cachette Project relies on the object and museum data described in the version 1 of the related database and on the photographic material added in its version 2. In order to fully use this already existing material, as well as to store the new data created throughout the encoding of the texts, Xefee leans on a MySQL database in which these different kinds of data are merged. Organised around a main “document” table, the data is spread over eighteen tables, among which four are dedicated to data from version 1, and one to the encoded texts. In order to make full use of this material in a XML perspective, a sixth and last tab of the GUI is dedicated to the creation of the XML/TEI files. By pressing the upper-left button, the user asks Xefee to pick up in the MySQL database all the needed pieces of information and to place them between the appropriate XML tags. This generates all the sections of a XML file, from the headers with the publication and bibliographic statements to the div elements dealing with the encoded texts. The newly created XML file will be then poured into a native XML eXist database in order to constitute the electronic corpus itself.