# Chatsistant MCP MCP

> Chatsistant connects your chatbots to any AI client, letting you manage the entire conversation flow through natural dialogue. You can list all bots, review every chat session, and even train them by adding new knowledge sources—all without logging into a dashboard. It gives developers and ops teams total control over their chatbot ecosystem from where they already work.

## Overview
- **Category:** customer-support
- **Price:** Free
- **Tags:** ai-assistant, white-label, conversation-analytics, knowledge-base, webhook-integration, bot-management

## Description

Managing customer-facing chatbots used to mean jumping between dashboards, digging through logs, or manually updating data feeds. Now, you treat your entire bot deployment like another service managed by conversation. You can list all the bots running and inspect exactly what they know, seeing their knowledge bases in real time. Need to debug a tricky interaction? Just review any chat session's full message history, no matter which bot was involved. If a bot needs more info, you don't upload files via a web portal; your agent simply adds the new data source for you. This ability makes it simple for developers to configure bots and monitor complex integrations across different channels using Vinkius' catalog of tools.

## Tools

### add_data_source
Adds a new URL, document, or text snippet to train a specific chatbot.

### get_bot
Retrieves the detailed profile and status for one particular bot.

### get_conversation
Fetches all messages and details from a single, specific chat session.

### list_bots
Provides an overview listing every chatbot currently configured in the system.

### list_conversations
Lists recent conversations, with an option to filter by which bot was involved.

### list_data_sources
Shows all the external data sources currently feeding information into a specific bot.

### list_webhooks
Lists every webhook setup, detailing its event triggers and delivery settings.

### query_bot
Sends a question to the bot and gets an answer based on its current knowledge base.

## Prompt Examples

**Prompt:** 
```
List all my bots and query the support bot about return policies.
```

**Response:** 
```
You have 3 bots: 'Support Agent' (bot_101, active), 'Sales Helper' (bot_102, active), and 'Internal Wiki' (bot_103, paused). Querying 'Support Agent' about return policies... The bot responded: 'Our return policy allows returns within 30 days of purchase. Items must be in original packaging with proof of purchase. Refunds are processed within 5-7 business days.' Would you like to review the data sources training this bot?
```

**Prompt:** 
```
Show recent conversations for the Sales Helper bot from this week.
```

**Response:** 
```
Found 23 conversations for bot_102 this week. The most active topics were 'pricing plans' (8 conversations), 'integration setup' (6 conversations), and 'enterprise features' (4 conversations). The longest session was conv_8891 (12 messages, 8 minutes). 3 conversations ended with a demo booking request. Would you like to inspect a specific session?
```

**Prompt:** 
```
Add our FAQ page and API documentation to the Internal Wiki bot.
```

**Response:** 
```
Adding 2 data sources to 'Internal Wiki' (bot_103)... ✅ URL 'https://docs.example.com/faq' added successfully (type: url, status: processing). ✅ URL 'https://docs.example.com/api' added successfully (type: url, status: processing). The bot now has 9 data sources total: 5 URLs, 3 documents, 1 text snippet. Processing will complete within 3-5 minutes. Would you like to unpause the bot?
```

## Capabilities

### Discover all deployed chatbots
List every configured bot, checking its status and knowledge base settings.

### Inspect full chat transcripts
Retrieve the complete message history for any specific conversation across any connected bot.

### Update a bot's training data
Programmatically add new URLs, documents, or text snippets to expand what a bot knows.

### Send questions to a bot's knowledge base
Query any active bot and receive an answer generated specifically from its trained data sources.

### Monitor webhook activity
View all configured webhooks, checking which event triggers them and how they send data.

## Use Cases

### Diagnosing a Bad Answer
A customer support agent sees a bot gave the wrong return policy. They ask their agent, 'Show me all conversations for the Support Bot last week.' The agent uses `list_conversations` to find the bad chat and then checks the data sources using `list_data_sources` to see if the correct policy was ever added.

### Updating Policies
The marketing team just launched a new pricing page. Instead of logging into three different systems, they ask their agent, 'Add the FAQ and API docs URL to the Internal Wiki bot.' The agent executes `add_data_source` instantly.

### Checking System Health
The ops team suspects a payment webhook is broken. They immediately run the check with `list_webhooks`. If it shows no recent activity, they know exactly where to focus their troubleshooting efforts.

### Initial Bot Audit
A new client signs up and needs an audit of all existing bots. They ask their agent to 'List every bot I run.' The agent uses `list_bots` to give them a clean, comprehensive overview.

## Benefits

- Stop hunting through dashboards. You can list all bots and use `get_bot` to check specific configurations, all without clicking a single button.
- Deep dive into every interaction using the MCP. `list_conversations` lets you pull full chat histories so you can spot exactly where the bot failed or succeeded.
- Never manually update knowledge bases again. You use `add_data_source` to programmatically feed new documentation (URLs, files) directly to a bot's memory.
- Testing answers is simple. Use `query_bot` to send questions and get instant, factual responses based only on the data sources you provided.
- Full operational visibility means knowing if your integrations are live. Run `list_webhooks` to verify every event trigger fires correctly.

## How It Works

The bottom line is you get total control over complex bot management without touching a graphical user interface.

1. Subscribe to this MCP and plug in your Chatsistant API Key from the dashboard.
2. Connect it via your preferred AI client (like Claude or Cursor).
3. Start managing bots, reviewing conversations, and adding data sources using natural language prompts.

## Frequently Asked Questions

**How do I list all my bots using the Chatsistant MCP?**
You use `list_bots` to get an overview of every bot. This tool returns a simple, actionable list showing the names and basic status of all your deployed chatbots.

**Can I check old chats with list_conversations? Is it hard?**
No, you just use `list_conversations` to see recent history. You can even narrow down the search by telling it which specific bot ID was involved.

**If I need to train a bot on new PDFs, do I use add_data_source?**
Yes, `add_data_source` is the tool you need. It lets you programmatically feed the system any type of data—URLs, files, or plain text—to expand the bot's knowledge.

**How do I know if my webhooks are working?**
Run `list_webhooks`. This tool gives you a clear rundown of every configured webhook, showing exactly what event is supposed to trigger it and where the data should go.

**What does query_bot actually do when I use it?**
It sends your question to a specific bot's knowledge base. The bot processes that question against its stored documents and returns an answer derived only from those sources.

**When I use get_bot, what specific configuration details can I pull for a single chatbot?**
You get detailed status information and settings for that bot. This includes its unique ID, whether it's currently active or paused, and direct access to the names of all data sources attached to it.

**If I use get_conversation, what kind of metrics can I pull from a specific chat session?**
You retrieve the full message history, allowing you to review who said what and when. It also provides key metadata about the conversation, like its total length or duration.

**How do I audit all my current data sources using list_data_sources?**
This tool lists every knowledge base attached to your bots. You can check which source is associated with which bot and see the current status of that data set.

**Can I send a question to a bot and get an AI-generated answer in real time?**
Yes! The `query_bot` tool accepts a Bot ID and a question string. It sends the query to the bot's AI engine and returns a response generated from its trained knowledge base — perfect for testing bot accuracy before deploying changes.

**Can I review all the data sources currently training my bot?**
Yes. The `list_data_sources` tool returns all URLs, documents, and text snippets that have been added to a specific bot's knowledge base, including their processing status. Use `add_data_source` to programmatically add new URLs, text, or file content to expand the bot's training data.

**Can I browse conversation histories across all my bots?**
Yes. Use `list_conversations` to retrieve all chat sessions — optionally filter by a specific Bot ID. Then use `get_conversation` with the Conversation ID to inspect the full message timeline, including user questions, bot responses, and timestamps.