User guide
User guide
Find guidance from choosing an archive and creating your first record through organization, AI connections and backups. If you are new, start with Getting started.
For first-time users
- Choose your environment and storage in Getting started.
- Read Record types and create your first diary and wiki records.
- Learn to collect records about the same person, project or topic in Organization and links.
An AI connection is not required to write or find records. Open the web app without installing anything; PC-local archives are available in the Windows app.
Browse by purpose
- Choose an archive, create records, manage membership: Getting started
- Events, schedules, goals and reference documents: Record types
- Period, relationships, projects, resources, topics and counts: Organization and links
- ChatGPT, Claude, server MCP and PC-local MCP: AI connections
- Sign-in, invitations, permissions, storage, backups and restoration: Troubleshooting
Use each guide’s contents to jump to a section. On narrow screens, expand Guide contents and On this page.
Things to know first
- Local and server storage are separate. Records do not automatically move between browser, PC and server archives. Signing in does not upload local materials.
- Classification requires real connections. Writing a person’s or project’s name in the body does not create group membership.
- File features are temporarily suspended. New file additions, opening and previews are unavailable. Existing files and links are retained and included in full backups. Markdown text documents remain available.
- You approve AI connections yourself. Content searched or read by AI is sent to the connected external service. Read what is transferred and how to disconnect first.
If you need help
If the guides do not resolve the issue, report a bug, suggest a feature or ask a usage question in Feedback. Sign in with your existing Alfred membership account. Only the author and designated operators can read messages and replies.
Do not paste entire archives or authentication tokens into inquiries. Share the problem, environment, steps taken and displayed error only to the extent needed. Feedback does not automatically collect archive materials or connect them to AI.