Layer 4 — MeshDB (Search)

Distributed full-text search engine for server meshes. SQLite FTS5 with optional semantic search via ChromaDB.

Install

pip install meshpop-db

What It Does

MeshDB indexes files across all your servers and provides fast full-text and semantic search. It gives AI the context to understand your entire codebase, configs, and documentation across the mesh.

CLI

meshdb index /path/to/project        # Index a directory
meshdb search "nginx config"         # Full-text search
meshdb search --smart "how auth works" # AI-enhanced search
meshdb find "docker-compose*"        # Find files by name
meshdb read /path/to/file            # Read indexed file
meshdb status                        # Index statistics
meshdb reindex /path                 # Re-index directory

Options: --json, --lang ko, --type py, --limit 20, --server myserver

MCP Tools

Tool	Parameters	Description
`meshdb_search`	`query` (required), `servers`, `limit`, `type`, `dir`	Full-text search across all servers
`meshdb_find`	`query` (required), `servers`, `limit`	Find files by name pattern
`meshdb_read`	`path`, `server` (required)	Read file contents from index
`meshdb_status`	`server`	Index stats — file count, size, last updated
`meshdb_semantic`	`query` (required), `limit`	Semantic search via AI embeddings

How It Works

Indexing: Scans directories, extracts text from files, stores metadata and tokens in SQLite FTS5
Search: BM25 ranking with snippet highlighting and language detection
Distributed: Each server maintains its own local index. Queries fan out across the mesh via vssh
Semantic (optional): ChromaDB vector embeddings for meaning-based code search

Interfaces

Interface	Usage
CLI	`meshdb search <query>`
Python	`import meshdb; meshdb.api_search(query)`
MCP	`meshdb-mcp` for AI tools

Source

meshpop/meshdb — pip install meshpop-db