Integration · Databases
AI agents for turbopuffer
Search, add, change and delete documents in turbopuffer namespaces: vector search, full-text (BM25) search, filters, ordering and counts, plus the list of namespaces and each one's size and schema.
- Actions
- 7
- Read
- 3
- Write
- 4
Actions
What your agents can do in turbopuffer
Search, add, change and delete documents in turbopuffer namespaces: vector search, full-text (BM25) search, filters, ordering and counts, plus the list of namespaces and each one's size and schema. One connection works with one turbopuffer region.
Read 3
Look things up. Nothing changes.
List namespaces
Lists namespaces, as namespaces (each with its id), 100 a page by default.
Get namespace details
Returns one namespace's schema (each attribute's type, and which are full-text searchable or hold vectors), approx_row_count, approx_logical_bytes, created_at, last_write_at and index.
Search documents
Queries one namespace and returns rows: each with id, the include_attributes asked for, and $dist (the score).
Write 4
Create, change or send.
Add or replace documents
Writes up to 1,000 documents to a namespace, creating it on the first write, as rows_upserted.
Change documents
Sets some attributes on existing documents, leaving the others as they are, as rows_patched.
Delete documentsDeletes data
Permanently deletes documents from a namespace, by ids or by filter (every match, same form as Search documents' filters, like ["status", "Eq", "archived"]), as rows_deleted.
Delete a namespaceDeletes data
Permanently deletes a whole namespace and every document in it.
A person can approve first
Any write action can be set to wait for a person's approval before it runs, so an agent asks before it changes anything. Actions that delete data ask first by default.
Connection
How it connects
Connect it once in the console, then give it to the agents and workflows that need it.
Paste an API key from your turbopuffer account.
What you enter
- API key
Settings
- Region
Kept safe
- Credentials are stored encrypted with your organization's own keys.
- They are opened only by the process that makes the call, at the moment it runs, and never go into a prompt or the browser.
- Calls go only to the tool's own servers.
Ideas
Ideas to start with
Starting points made only from the actions above. Mix them with any other tool your agents can use.
- 01
Ask about your turbopuffer data
Ask a question in plain words. The agent can list namespaces, get namespace details or search documents in turbopuffer, then answers with what it found.
- 02
Let an agent update turbopuffer
Have the agent add or replace documents or change documents in turbopuffer as part of a larger task. Turn on approval and a person signs off on each change first.
- 03
Share what it finds in Slack
Give one agent turbopuffer and Slack: it can list namespaces in turbopuffer and post a summary to a Slack channel in the same run.
Works well with
More Databases tools
- PostgreSQL5 actionsRead and query a PostgreSQL database directly: Amazon RDS and Aurora, Google Cloud SQL, Azure, Neon, Supabase, or your own server with a public address.
- MySQL5 actionsRead and query a MySQL database directly: Amazon RDS and Aurora MySQL, Google Cloud SQL, Azure, PlanetScale, or your own server with a public address.
- MongoDB7 actionsRead and query a MongoDB database directly: MongoDB Atlas (mongodb+srv) or your own server with a public address.
- BigQuery5 actionsRun SQL on BigQuery, browse datasets, tables and their columns, and stream rows into a table, in a Google Cloud project, signed in with the person's own Google account.
- Snowflake9 actionsRun read-only SQL in Snowflake and page through the results, and list databases, schemas, tables, columns and warehouses, through the SQL API with a programmatic access token or a key pair.
- Pinecone8 actionsSearch, add, read and delete vectors in a Pinecone index: similarity search by vector or by a stored record's ID, text search on indexes with integrated embedding, index statistics and the list of indexes in the project.
FAQ
turbopuffer and MyDeepAgents
Can MyDeepAgents connect to turbopuffer?
Yes. The turbopuffer connector has 7 actions: 3 read and 4 make changes. Your agents can call them as tools, and your workflows can run them as steps.
How do I connect turbopuffer?
Paste an API key from your turbopuffer account. You'll need: API key. The connection also asks for: Region. Credentials are stored encrypted with your organization's own keys and are opened only at the moment of a call.
Can an agent change data in turbopuffer without asking?
Only if you let it. Each of the 4 actions that change data can be set to wait for a person's approval before it runs, and the 2 that delete data ask first by default.
Can I use turbopuffer in workflows as well as agents?
Yes. Every turbopuffer action also works as a step in a workflow, and the same connection serves your agents and your workflows.
Early access
Put turbopuffer to work
Connect it in a few minutes, give an agent the actions it needs, and decide which changes wait for a person.