Meilisearch MCP server

What the model can do

Meilisearch (self-hosted or Meilisearch Cloud): list indexes, their statistics and settings, search one index or several, search facet values, read one document, see the indexing tasks that ran, failed or are queued, and, when asked, replace, update or delete documents (one, a listed batch or by filter), empty an index, create, delete or swap indexes, change an index's settings, synonyms or primary key, cancel or delete listed tasks, and start a dump or snapshot.

Reads

• meilisearch_list_indexesMeilisearch indexes: uid, primary key, created and updated times, and the total.
• meilisearch_index_statsStatistics for one Meilisearch index: number of documents, whether it is indexing right now, and how many documents carry each field…
• meilisearch_searchSearch one Meilisearch index and return the matching documents, each cut to 30 fields.
• meilisearch_get_documentOne document from a Meilisearch index by its primary key value, cut to 30 fields.
• meilisearch_list_tasksMeilisearch's task queue, newest first: document additions, deletions, settings changes and index operations, with status (enqueued, processing…
• meilisearch_get_settingsThe settings of one Meilisearch index: displayed, searchable, filterable and sortable attributes, ranking rules, stop words, synonyms…
• meilisearch_multi_searchSearch several Meilisearch indexes (or one index several ways) in one request; each query's hits come back separately, each document cut to 30 fields.
• meilisearch_facet_searchThe values of one facet (a filterable attribute) in a Meilisearch index with how many documents carry each…

Acts, when you ask

These change something in Meilisearch. The model is told to use them only when you ask for exactly that, and each one is granted by name: per team on a shared connection, by the owner on a personal one.

• meilisearch_replace_documentsAdd documents to an existing Meilisearch index.
• meilisearch_update_documentsUpdate documents in an existing Meilisearch index.
• meilisearch_delete_documentPermanently delete one document from a Meilisearch index by its primary key value.
• meilisearch_delete_documents_by_filterPermanently delete every document in a Meilisearch index that matches a filter (attributes in the filter must be filterable).
• meilisearch_delete_all_documentsEMPTY a Meilisearch index: permanently delete every document in it (settings and the index itself are kept).
• meilisearch_create_indexCreate a new, empty Meilisearch index, optionally naming its primary key.
• meilisearch_delete_indexPermanently DELETE a whole Meilisearch index: every document and all its settings.
• meilisearch_delete_documents_batchPermanently delete the listed documents (up to 100, by primary key value) from a Meilisearch index.
• meilisearch_update_settingsChange settings of a Meilisearch index.
• meilisearch_update_synonymsREPLACE the whole synonym list of a Meilisearch index with the one given (synonyms not in it are removed; {} clears them all).
• meilisearch_update_indexChange the primary key of a Meilisearch index.
• meilisearch_swap_indexesAtomically swap two existing Meilisearch indexes: their documents, settings and task history trade names…
• meilisearch_cancel_tasksCancel the listed Meilisearch tasks (by uid) that are still enqueued or processing; a cancelled document addition or settings change is not applied.
• meilisearch_delete_tasksPermanently delete the listed finished tasks (by uid) from Meilisearch's task history.
• meilisearch_create_dumpStart a dump of the whole Meilisearch instance (every index's documents and settings…
• meilisearch_create_snapshotStart a snapshot of the Meilisearch instance (an exact copy of its database in the server's snapshot directory) as a backup.

Connect it

First connect Meilisearch on Dashboard → Connections, then add Memtro to your client once. The Meilisearch tools appear alongside the tools of everything else you have connected; nothing is configured per connector.

claude plugin marketplace add memtro-org/claude-memtro
claude plugin install memtro@memtro --config api_key=sk-memtro-...

{
  "mcpServers": {
    "memtro": {
      "url": "https://platform.memtro.com/api/mcp",
      "headers": {
        "Authorization": "Bearer sk-memtro-..."
      }
    }
  }
}

{
  "mcp": {
    "memtro": {
      "type": "remote",
      "url": "https://platform.memtro.com/api/mcp",
      "headers": {
        "Authorization": "Bearer sk-memtro-..."
      }
    }
  }
}

Who may use it

The key identifies the person, and the server offers that person the tools they may use everywhere else in Memtro. A shared Meilisearch connection is used by the teams an administrator granted it to, tool by tool, on the Permissions page; a personal connection is used by its owner alone, and its tools that act stay off until the owner allows them. See Teams and permissions and Connectors.

Ask it like this

• › Did last night's import into the products index fail?
• › Why does a search for blue kettle return nothing?
• › How many documents are in the articles index?
• › Delete the document with id 4471 from the products index
• › Add kettle as a synonym for jug in the products index

Questions

Is this an official Meilisearch MCP server?

No. It is Memtro's MCP server with Memtro's Meilisearch connector behind it: the connector reads Meilisearch through its API with the credentials you connect on Dashboard → Connections, and Meilisearch publishes nothing here. The same endpoint carries every other system you connect and Memtro's memory tools.

What does it cost?

Nothing on top of Memtro. One price per organisation by its number of users, from $50 a month for up to 10 users after a 7-day free trial, with every connector, the MCP server, jobs and memory included. You bring your own model keys.

Can it change anything in Meilisearch?

Only when you ask, and only what it has been granted: add documents to an index, replacing whole any document with the same primary key; update fields of documents in an index, adding any that do not exist; delete one document from an index; delete the documents in an index that match a filter; empty an index (delete every document); create an index; delete an index with all its documents and settings; delete up to 100 listed documents by id; change an index's settings (searchable, filterable, sortable and displayed attributes, ranking, stop words, synonyms, typo tolerance, limits); replace an index's synonyms; change an index's primary key; swap two indexes; cancel listed queued or running tasks; delete listed tasks from the task history; create a dump of the instance; create a snapshot of the instance. Every other tool reads, searches or lists. An administrator grants each acting tool by name per team, and a personal connection acts only after its owner allows it.

Which clients work?

Any MCP client that speaks Streamable HTTP: Claude Code, Cursor, Claude Desktop, OpenCode, Codex, Zed, Windsurf, Cline and the Anthropic and OpenAI SDKs. There is no local process to run; the server is remote and authenticated with your key.

Other MCP servers · Databases and warehouses

• Postgres MCP server
• MySQL and MariaDB MCP server
• SQL Server MCP server
• BigQuery MCP server
• Snowflake MCP server
• AWS Athena MCP server