gid
A personal research agent
gid answers questions from a memory it keeps. When the memory is thin or old, it goes to the web through gates: a search, trusted feeds, an archive of primary documents. Whatever it finds is stored with where it came from: the document, the publisher, the date it was read, and how far the claim is from its source.
This page is the front door. Behind it is the interface to that memory: private and read-only, opened with a password.
What we built
- System: background jobs, failure markers, backups, and how much memory the host has left.
- Search: the stored records, found by their words and narrowed by tags.
- Memory: what the memory holds and what it rests on: primary documents or retellings, sourced numbers or unexplained ones.
- Reminders: open questions, each with the date of its next check.
- Documents: the reference pages and the records each of them stands on.
- Day plan: today's part of the work plan.
- Inbox: a page where files go from a computer to the host, for gid's sessions to read.
Sign in
- Open the interface The browser asks for the login and password.
- Upload files Opens only through the SSH tunnel to the host.
If “gid” is a word, it probably is short for “guide”. It is short for “general info dump”, which is what it was before it learned to keep its sources.