pyvectorhound.toml

[mcp]
enabled = true
port = 8782
discovery_method = "stdio"

[mcp.tools]
create_vector_index = true
index_documents = true
semantic_search = true
similarity_search = true
hybrid_search = true
retrieve_by_id = true
batch_retrieval = true
cluster_documents = true
rerank_results = true
get_index_stats = true
delete_from_index = true
update_documents = true
export_index = true

[mcp.security]
require_auth = false