Crate toondb_grpc 

ToonDB gRPC Services

This crate provides a comprehensive gRPC interface for ToonDB operations. It implements a “Thick Server / Thin Client” architecture where all business logic lives in the Rust server, enabling thin SDK wrappers in any language.

Services

• VectorIndexService: HNSW vector operations
• GraphService: Graph overlay for agent memory
• PolicyService: Policy evaluation and enforcement
• ContextService: LLM context assembly with token budgets
• CollectionService: Collection management
• NamespaceService: Multi-tenant namespace management
• SemanticCacheService: Semantic caching for LLM queries
• TraceService: Trace/span management
• CheckpointService: State checkpoint and restore
• McpService: MCP tool routing
• KvService: Basic key-value operations

Usage

# Start the gRPC server
toondb-grpc-server --port 50051

# From Python client
import grpc
from toondb.proto import toondb_pb2, toondb_pb2_grpc

channel = grpc.insecure_channel('localhost:50051')
stub = toondb_pb2_grpc.VectorIndexServiceStub(channel)

Re-exports

pub use server::VectorIndexServer;

pub use error::GrpcError;

Modules

checkpoint_server

Checkpoint Service gRPC Implementation

collection_server

Collection Service gRPC Implementation

context_server

Context Service gRPC Implementation

error

Error types for gRPC service

graph_server

Graph Service gRPC Implementation

kv_server

KV Service gRPC Implementation

mcp_server

MCP Service gRPC Implementation

namespace_server

Namespace Service gRPC Implementation

policy_server

Policy Service gRPC Implementation

proto

semantic_cache_server

Semantic Cache Service gRPC Implementation

server

gRPC Server implementation for Vector Index Service

trace_server

Trace Service gRPC Implementation