Recall
The Grepp agent can look back through your conversation to find past queries, results, and work — so you don't have to repeat yourself or dig through history manually.
What the agent remembers
Throughout a conversation, Grepp automatically tracks:
- Queries run — the SQL behind every chart or result
- Charts and dashboards created or modified — their titles, IDs, and configurations
- Filters and parameters you've applied — date ranges, groupings, conditions
- Your questions and requests — the context of what you were trying to accomplish
This session-level memory means the agent always knows what it built and can reference it without re-running queries.
Asking the agent to recall
You can ask the agent to look back at any point in the conversation:
- "What was the query we used for the sales breakdown?"
- "Can you show me that chart we built earlier?"
- "What tables have we queried this session?"
- "Summarize what we've done so far"
The agent will search through conversation history and surface the relevant information.
Reading dashboards
The agent can also retrieve the current state of any dashboard — not just what was built this session. This is useful when you want to reference or modify an existing chart:
- "What charts are on my Marketing dashboard?"
- "Update the revenue chart on the Executive Overview dashboard"
The agent fetches the current dashboard state and uses it to respond accurately.
The agent prioritizes what was built in the current conversation when you reference recent work. For older work from previous sessions, you can ask it to check a specific dashboard by name.