Sacred Knowledge Vault

An AI knowledge tool with an exclusive role-playing system featuring two separate knowledge base modes. All conversation data is stored locally to fully protect your privacy.  Six Exclusive Roles： You must pick an identity on your first launch: Angel, Demon, Sage, Knight, Fox and Hermit. Each role comes with unique salutations, conversation tones and placeholder texts for the input box. You may switch roles anytime in Settings. The AI opens every reply with role-specific greetings for an immersive Q&A experience.  Dual Knowledge Base Modes — Balance of Convenience & Privacy： 1. Sacred Library — Official Built-in Public Knowledge Base It includes ten complete categories: full-stack development, general learning, quotes & proverbs, design thinking, workplace communication, history & literature, finance & psychology, space science and more. Pre-built prompt templates named Wisdom Sparks are available. You can read daily offline quotes without API configuration — ready to use right after installation.  2. Personal Codex — Custom Private Model Knowledge Base (Max Privacy Protection) Create your private knowledge vault by entering your LLM API address, secret key and model name. All conversations are only sent to the third-party interface you configure. The app will not collect, upload or retain your Q&A content on its own. Full chat history is saved locally in the on-device FMDB database with no cloud uploads. You retain full ownership of all your data.  Full Suite of Knowledge Management Tools： - Sacred Archives: Conversations are auto-saved locally, paginated and grouped by knowledge base; only accessible on your device. - Sacred Favorites: Long-press any dialogue to save it locally for later review. - Wisdom Sparks: Built-in prompt templates to start knowledge inquiries with one tap.  Privacy & Security Guarantees： 1. Local Data Storage: All chats, favorites and history remain in your local database with no automatic cloud uploads. 2. Isolated Private Mode: Messages in your private vault are only relayed to your self-provided model endpoint. We never access your API keys or conversation data. 3. Encrypted API Key Storage: Your custom model keys are securely stored in Keychain with no plaintext caching. 4. Full Data Control: Clear all local chat cache and device data in one tap via Settings.  Target Users： Programmers, designers, educators, copywriters, students, casual knowledge seekers, and users who run private LLMs and prioritize data privacy. We keep updating built-in knowledge bases and role interaction experiences continuously.