For extracting cover pictures from uploaded books the library "Imagemagick" has to be installed. The project website lists how to install binary releases for OS-X and windows: https://imagemagick.org/script/download.php. For Linux it's recommended to use the imagamagick version of your distro. (For Debian/Ubuntu/Linux Mint: sudo apt install imagemagick).
Extracting metadata from fb2 or epub files needs additional dependencies from optional-requirements.txt. In the optional-requirements.txt there are different blocks for the different options. Install the relevant requirements via pip/pip3.
Use your own robots.txt file
To make Calibre-Web deliver a robots.txt file, place the file with the relevant content in the cps/static subfolder of Calibre-Web. The filname has to be robots.txt. The file is persitent even over upgrading Calibre-Web via the user interface.