Skip to content

lorehq/lore-os-services

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

lore-os-services

Docker services stack for Lore OS. Provides semantic search over the DATABANK and scoped hot memory backed by Redis.

What's Inside

A single container runs three processes:

  • Redis -- persistent hot cache for session, project, and global memory
  • FastAPI -- semantic search and memory API (port 8080)
  • Watcher -- monitors the DATABANK for file changes and triggers incremental reindexing

The embedding model (BAAI/bge-small-en-v1.5) is baked into the image at build time. No network access is required at runtime.

Usage

The container is managed by the Lore OS bundle. The docker-compose.yml lives in the lore-os repo (~/.lore-os/docker-compose.yml) and is started/stopped from there.

Pull and run manually:

docker pull lorehq/lore-os-services:latest
docker run -d \
  -p 9184:8080 \
  -v ~/LORE-OS/DATABANK:/data/DATABANK:ro \
  -v ~/LORE-OS/HOT:/data/redis \
  lorehq/lore-os-services:latest

Volumes

Mount Container path Mode Purpose
~/LORE-OS/DATABANK /data/DATABANK ro Markdown files indexed for semantic search
~/LORE-OS/HOT /data/redis rw Redis AOF persistence

API

All endpoints are served on port 8080 inside the container.

Method Path Description
GET /health Service status, index stats
GET /search?q=...&k=10 Semantic search over DATABANK
POST /reindex Full reindex of all markdown files
POST /reindex-file Incremental reindex of a single file
POST /activity Record file access for hot-path tracking
POST /memory/hot/write Write scoped memory (session/project/global)
GET /memory/hot/recall Recall scoped memory entries
GET /memory/hot/stats Memory entry counts by scope

Set LORE_TOKEN to require Bearer token auth on all requests.

Environment

Variable Default Description
DOCS_SOURCE /data/DATABANK Root directory for markdown indexing
SEMANTIC_PORT 8080 FastAPI listen port
EMBED_MODEL BAAI/bge-small-en-v1.5 Embedding model (must be baked in)
DEFAULT_K 10 Default number of search results
MAX_K 10 Maximum allowed k
RESULT_MODE paths_min Default search output format
LORE_TOKEN (none) Bearer token for API auth
REDIS_URL redis://localhost:6379 Redis connection URL

Building

docker build -t lorehq/lore-os-services:latest .

License

Apache 2.0

About

Lore memory engine — semantic search + hot memory sidecar

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors