other
Weaviate is a vector database, and getting its indexing and query syntax right usually means jumping between docs and a running cluster. The Weaviate Docs MCP Server puts the documentation itself inside the conversation, so Neotask's agent can look up how a class schema, a hybrid search query, or a vector index setting actually works while it's building or debugging against your Weaviate instance. Instead of you copy-pasting error messages into a search bar, the agent reads the current docs, checks them against what you're trying to do, and applies the answer directly to your data or your code.
| Weaviate Docs MCP Server | Look up Weaviate documentation and query a Weaviate vector database for storing and searching objects, vectors, and indexes used in AI and search applications. |
Tell the agent your hybrid search query isn't ranking the way you expect. It checks the Weaviate docs for how alpha weighting and BM25 combine with vector distance, compares that against your actual query against the connected cluster, and adjusts the query or schema accordingly.
Describe the objects you want to store — say, product listings with embeddings and a few filterable properties. The agent looks up the current schema syntax in the docs and creates the class directly in your Weaviate instance, rather than guessing at deprecated field names.
It's most useful when you have a Weaviate cluster to query, but the documentation lookup half works on its own — the agent can explain concepts or syntax even before you've created a database.
Yes — it can store objects, vectors, and indexes, not just query existing ones, as long as the connection has access to your cluster.