Every interaction with your agent creates a conversation — a thread of messages between a user and the agent. You can review conversations to understand what users are asking and how well your agent is performing.Documentation Index
Fetch the complete documentation index at: https://mind-e.co/docs/llms.txt
Use this file to discover all available pages before exploring further.
Viewing chat history
Go to your agent’s Chat History tab to see all conversations. Each conversation shows:- Source — where the conversation happened (widget, Telegram, WhatsApp, Instagram, Messenger)
- Title — auto-generated from the conversation content
- Date — when the conversation started
- Message count — number of messages exchanged
Message details
Each message in a conversation includes:- Role — whether it’s from the user or the agent
- Content — the message text
- Sources — for agent responses, the documents or snippets that were used to generate the answer, along with relevance scores
- Images — any images sent by the user
- Timestamp — when the message was sent
Source attribution
When the agent answers a question using content from the knowledge base, the response includes sources — references to the specific document chunks that were used. Each source shows:- The document name
- The relevant text excerpt
- A relevance score
Conversation sources
Conversations are categorized by where they originated:| Source | Description |
|---|---|
| Playground | Testing in the dashboard |
| Widget | Embedded chat widget on your website |
| Instagram Direct Messages | |
| WhatsApp Business messages | |
| Telegram | Telegram bot messages |
| Messenger | Facebook Messenger messages |
Analytics
Go to the Analytics tab to see aggregate metrics for your agent:- Total conversations and messages
- Conversations over time
- Credit usage breakdown
- Source distribution (which channels generate the most conversations)
Tips
- Review regularly — check conversations to spot common questions your agent can’t answer, then add that content to the knowledge base
- Check sources — if the agent gives incorrect answers, check which sources it’s using and update the relevant documents
- Monitor credit usage — analytics show how many credits each agent uses, helping you optimize costs