This agent generates a vector embedding using OpenAI's ada-002 text embedding model. The output can then be used by other agents to query external vector databases which use the same embedding model on Pinecone, Supabase, etc. 

Required inputs for this agent include a simple text string input and an OpenAI API key for retrieving the vector embedding via API. 

Vector Embedding Generator

This agent will parse out JSON items from a full JSON string, based on a predefined position for the target object and/or a specific header. This agent is best used as a part of other complex automations on agent.ai.

As an example, if another agent is returning a JSON string with multiple objects and/or headers, but you only want to utilize the 3rd item in your automation, and the header "name", provide the original JSON string, the number 3 for "target_object", and "name" for "target_header" . The output will contain only the name from the 3rd JSON item, with it's original header, in a usable JSON format. 

Connor Davidson

Associated Agents

Vector Embedding Generator

JSON Object Parser