The aim of this paper is to present an application to help speech therapists characterize the lexical properties of the written texts of patients with dyslexia and dysorthographia. Recent studies in France about written production of students with dyslexia (Authors, 2021, 2022) reveal: 1- that students with dyslexia use the same words; 2- but they make many more errors than control students; and 3- they produce an unexpected subset of errors concerning short words (particularly determiners and prepositions) and specifically words that are both short and homophonic. For these previous studies, a lexical processing chain has been developed to assign various linguistic information to each word used by students. Based on the Lexiques (Lexique.org) and Manulex (Lété, Sprenger-Charolles & Colé, 2004) tools, we can indicate for each word: 1- grapheme and phonetic properties; 2- Syntactic properties; 3- frequency properties and 4- length properties. To help specialists in written language rehabilitation, these results and the lexical processing chain are currently being exploited to meet the needs of speech therapists to better specify spelling errors. Based on these previous studies and on the lexical processing chain, we developed a first version. We are currently testing it with a panel of speech therapists and the first clinical test results will be presented during this presentation. This step of testing will enable us to adapt the tool to the needs of speech therapists to better specify the spelling errors of people with dyslexia and thus better guide remediation. This tool is an example of a successful interface between theoretical research and practical application, and an argument for encouraging transdisciplinary work.