agents
latest
false
- Getting started
- Prerequisites
- Building agents
Agents user guide
Last updated Mar 7, 2025
Contexts
Indexes contain reference documents specific to a business process or business unit. An agent may need to reference the knowledge stored in these indexes at any point while it runs.
To learn how to create an index, see Adding a new index in the Context Grounding documentation.
To add context for your agent, take the following steps in the Contexts section:
- Provide a Description for your context.
- Select the shared Orchestrator folder where your index is located.
- In the ECS Index field, select the index from the available list. The dropdown list includes all indexes available in the specified folder.
- Configure additional
settings. You can control the precision and relevance of the context that
you want to query from using the following settings:
- Threshold: Refers to a similarity score threshold applied after the retrieval and re-ranking of relevant chunks. Any relevant chunks that are scored below this threshold are not returned. Use Threshold to make results more precise. As you increase this number, Context Grounding filters and grounds prompts with only the best scores.
- Number of results: Refers to the number of chunks returned by Context Grounding from a user's query. It helps with making the grounding payload of the LLM generation (overall token size of the LLM call) smaller and can help act as an additional filter on the results you get from Context Grounding when Context Grounding is applied. Increasing the number of results increases the amount of chunks that are returned.
