> ## Documentation Index > Fetch the complete documentation index at: https://docs.presenton.ai/llms.txt > Use this file to discover all available pages before exploring further. # Presentation context and memory > Understand what the assistant can use and what you still need to state explicitly. Editing context can include the current presentation, selected slide, outline, source files, template, and recent conversation. Context improves continuity, but it is not a guarantee that every earlier detail will be preserved. ## How presentation context helps * Understands the overall presentation structure. * Maintains relevant context across slides. * Helps keep tone, style, and narrative flow consistent. * Uses existing deck content as a reference while editing. ## Work safely with context * Identify the exact slide or section in each important request. * Restate facts and constraints that must not change. * Point to the relevant attachment when a request depends on source evidence. * Review the saved presentation after every edit. * Treat the visible deck, not assumed model memory, as the record of the work. Cloud manages application storage and the hosted assistant workflow. Do not assume an unrelated presentation or a new conversation automatically carries earlier instructions. Repeat critical context in the request. Open-source Presenton can use presentation-scoped Mem0 memory when it is enabled and configured. Memory data is stored under the persistent app-data path by default and depends on its configured model and embedding services. Keep `/app_data` persistent. Replacing a container without its mounted app data can remove presentations, uploads, exports, settings, authentication state, and local memory data. Review persistence, database, memory, document parsing, and telemetry settings.