RAG Containers (Preview)
Planned — not available in version 1.0.0. A RAG container would package your exported conversations into a portable, self-contained knowledge base that other local tools can load without re-exporting anything.
What it will do
- A single file or folder holding your archive plus the embedding index, so it can be copied to another machine or another tool.
- Built from the Markdown you already have — no re-download, no provider access at build time.
- A documented, stable layout so an existing local LLM tool can point at it directly.
- Incremental rebuilds as new conversations land, rather than reconstructing from scratch.
- Everything stays local: no hosted vector database, no upload, no API key.
Why it is shaped this way
The point of a container is portability, and portability is mostly a contract: a fixed layout, a version number, and a manifest describing what is inside. Smart Search Engine (SSE) now ships with its own index, so the container has something settled to carry — the open question is which index format to standardise on and how to version it as the index evolves.
Want it sooner, or would your setup need something different? Tell us at contact@totalrecalls.app. Planned features are ordered by the requests that actually arrive, not by the order they appear on this page.