Hyperspell provides agents created with AgentStack with the ability to build and query internal knowledge using:
Create your Hyperspell app and get your API key from the Hyperspell Dashboard.
Follow the instructions to connect your data sources
Set the environment variables in your project’s .env
file:
Hyperspell is a multi-tenant platform, and you can separate your data by use by setting the user id of your end-user in a variable too:
The Hyperspell tool provides three core functions for knowledge retrieval and ingestion:
Search across your Hyperspell-connected sources for relevant information.
Parameters:
The search query to find relevant information
Comma-separated list of sources to search (e.g., collections,notion,gmail
)
If True, returns a direct AI answer instead of just documents. Defaults to False
User ID to use for this request. Defaults to HYPERSPELL_USER_ID
env var
Add a text document to your Hyperspell knowledge base.
Parameters:
The full text content to add
Optional title for the document
Optional collection name to organize the document
User ID to use for this request. Defaults to HYPERSPELL_USER_ID
env var
Upload a file (PDF, Word doc, spreadsheet, etc.) to your Hyperspell knowledge base.
Parameters:
Path to the file to upload
Optional collection name to organize the document
User ID to use for this request. Defaults to HYPERSPELL_USER_ID
env var
If no user_id
is provided, the tool defaults to the HYPERSPELL_USER_ID
environment variable.
Hyperspell can search across multiple integrated sources:
For detailed integration setup, visit the Hyperspell documentation.