search_contextLABSCO SUMMARY
Where it earns its place is the folder, not the snippet.
Plenty of servers give an assistant a place to remember things. This one is built around dropping a directory into the index — images, CAD, PDFs, data files and source all get their own metadata extraction — and quick_load makes that cheap enough to do for a folder you only care about for an hour.
What it is
An MCP server that keeps an assistant's memory in a ChromaDB instance running on your own hardware. Beyond storing snippets, it walks directories, extracts metadata per file type, and makes the result searchable — nothing leaves the machine.
What you get
- Memory tools —
store_context,search_context,list_collections,find_similar_patterns - Bulk ingest —
scan_directoryto preview what is there, thenbatch_ingest,ingest_file, orquick_loadfor a throwaway collection you drop again withunload_collection - Collection management —
get_collection_info,export_collectionandimport_collectionfor JSON backups,list_file_types - Photo metadata —
extract_exifpulls camera, lens, exposure, GPS and date taken out of JPEG and TIFF files - Watched folders —
watch_folder,stop_watch,list_watchersauto-ingest anything dropped in - Duplicate detection —
find_duplicateswith partial, full or perceptual hashing, pluscompare_filesandfind_collection_duplicates - Six collections created by the setup script, including code_snippets, configurations, troubleshooting and project_docs
Requirements
Bun as the runtime, Docker to run ChromaDB, and an MCP client. The documented setup runs the chromadb/chroma image with host port 8001 mapped and a persistent volume, then bun install and bun run setup to create the collections. Your client config points at index.js with CHROMADB_URL set to the ChromaDB address. MIT licensed.
Setup effort
Build from source — git clone https://github.com//vespo92/chromadblocal-mcp-server.git
USE CASES
01
Searching a photo dump you never tagged
Ingest the folder with EXIF extraction on, then ask by camera, place or date instead of scrolling thumbnails.
02
Reclaiming disk space across scattered directories
Hash-scan the tree, see the duplicate groups with the space they waste, and confirm any pair before deleting.
03
Loading a directory for one session and dropping it
Quick-load into a temporary collection, ask your questions, then unload it and free the memory.
20 tools
Listed in project docs. Your client may show a different list.
store_contextlist_collectionsfind_similar_patternsscan_directorybatch_ingestquick_loadunload_collectionexport_collectionimport_collectionget_collection_infoingest_filelist_file_typesextract_exifwatch_folderstop_watchlist_watchersfind_duplicatescompare_filesfind_collection_duplicatesBEFORE YOU INSTALL
What to know first
Tools you get
20 tools named in the project's documentation.
Where it runs
The documented client runs the server on your machine.
INSTALLATION
Connect your client
You clone this one yourself.
Server setup
TERMINAL
git clone https://github.com//vespo92/chromadblocal-mcp-server.git
cd chromadb-mcp-server
bun installTERMINAL
docker run -d \
--name chromadb-local \
-p 8001:8000 \
-v ~/chromadb-data:/chroma/chroma \
-e IS_PERSISTENT=TRUE \
chromadb/chroma:latestTERMINAL
bun run setupMCP Clients
Claude Desktop
CLAUDE DESKTOP · JSON
{
"mcpServers": {
"chromadb-context": {
"command": "bun",
"args": [
"run",
"<YOUR_PATH_TO_CHROMADB_MCP_SERVER>/index.js"
],
"env": {
"CHROMADB_URL": "http://localhost:8001"
}
}
}
}
