ChatFly MCP. Build and Deploy Highly Trained Chatbots.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
ChatFly connects your AI client to a full suite of chatbot management tools. It lets you build specialized conversational bots that answer product questions, qualify leads, and even book meetings without any human intervention.
You train these agents by feeding them external websites or documents, allowing them to act like expert support staff 24/7.
What your AI agents can do
Chat
Sends a message to an existing chatbot and gets its response.
Create bot
Creates a new, distinct chatbot instance with a custom welcome message.
Get bot
Retrieves the specific details and configuration of one existing bot.
Create, retrieve details for, and list all specialized chatbot bots.
Programmatically ingest data from websites or documents to update a bot's knowledge base.
Update existing bots, changing welcome messages or internal metadata without rebuilding them.
Send a message to any specific bot and retrieve its immediate AI response for testing or integration.
List all the URLs and documents currently feeding knowledge into your bots.
Ask AI about this MCP
Supported MCP Clients
OAuth 2.0 Compatible Gemini Waiting for input…
ChatFly MCP: Managing Your Bots with 7 Tools
Manage your entire bot lifecycle—from creation and training to real-time interaction—using this suite of seven specialized tools.
Make your AI actually useful.
Add this MCP to Claude, Cursor, or Windsurf and your AI stops guessing. It gets real tools to look things up, take action, and handle the stuff you keep doing by hand.
Start using ChatFly on Vinkius019dd0ccchat
Sends a message to an existing chatbot and gets its response.
019dd0cccreate bot
Creates a new, distinct chatbot instance with a custom welcome message.
019dd0ccget bot
Retrieves the specific details and configuration of one existing bot.
019dd0cclist bots
Provides a list of all chatbots currently managed by your account.
019dd0cclist data sources
Lists every URL or document source that is contributing knowledge to your bots.
019dd0ccupdate bot
Changes the settings, like welcome messages, for a bot you already built.
019dd0ccupload data source
Adds a new knowledge source (like a website URL) to train one or more bots.
Choose How to Get Started
Build a custom MCP for your own tools, or connect a ready-made integration from our catalog.
Build Your Own
Turn any API into an MCP. Import a spec, define Agent Skills, or deploy with MCPFusion.
- Import from OpenAPI, Swagger, or YAML specs
- Create Agent Skills with progressive disclosure
- Deploy to edge with MCPFusion framework
- Built in DLP, auth, and compliance on every call
- Real time usage dashboard and cost metering
- Publish to catalog or keep private
Make Your AI Do More
Start with ChatFly, then connect any of our 5,000+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 5,000+ others, all in one place
- Add new capabilities to your AI anytime you want
- Every connection is secured and compliant automatically
- Track usage and costs across all your servers
- Works with Claude, ChatGPT, Cursor, and more
- New servers added to the catalog every week
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by ChatFly. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.
VINKIUS INFRASTRUCTURE
Cloud Hosted
Managed infra
V8 Isolated
Sandboxed per request
Zero-Trust Proxy
No stored credentials
DLP Enforced
Policy on every call
GDPR Compliant
EU data residency
Token Compression
~60% cost reduction
Works with Claude, ChatGPT, Cursor, and more
The Model Context Protocol standardizes how applications expose capabilities to LLMs. Instead of operating in isolation, your AI gains direct access to external platforms, live data, and real-world actions through secure, standardized connections.
This server provides 7 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
The Manual Pain Point: Keeping Your Support Bots Current
Every time your product updates or a policy changes, someone has to manually go into the chatbot dashboard. They copy the new text from the internal wiki, find the right bot profile, and paste it in. This is tedious, error-prone work that takes hours of highly paid employee time.
With this MCP, you bypass that entire process. You just point your agent at a source URL—a single link to the updated documentation—and run `upload_data_source`. The system does the heavy lifting, ingesting and training the bot on the new data automatically. Now it's fast.
The Power of Bot Orchestration
Instead of relying on one massive 'catch-all' bot that struggles to distinguish between billing questions and technical failures, you can now use `create_bot` to build specialized agents. You set up a dedicated Lead Qualifier bot, separate from your Technical Support Bot.
This separation means each bot is highly focused. They know their job inside and out. Your AI client routes the conversation correctly, giving users a better experience and keeping your internal logic clean.
What you can do with this MCP connector
You're trying to automate customer interactions, but managing the knowledge base is a nightmare. This MCP lets your AI client take over that entire process. Instead of manually copy-pasting website content into training fields, you simply point the bot at URLs or documents using upload_data_source. The bot then builds a deep, accurate knowledge graph from that source material.
You can manage multiple specialized bots—say one for billing questions and another for product specs—all through one interface. Need to check how well a specific bot is performing? Your agent handles operational monitoring and session history retrieval automatically. This full capability set makes it easy to keep your support agents current, giving you total control over the knowledge base from anywhere within the Vinkius catalog.
019dd0cc-6aa3-7354-b7e6-04adf1f5c412 
How ChatFly MCP Works
- 1 First, you connect this MCP to your preferred AI client via Vinkius.
- 2 Next, use the agent to execute commands like
create_botorlist_botsto build and configure your bot network. - 3 Finally, tell the agent to run
upload_data_sourcepointing it at a URL, triggering the knowledge ingestion and training process.
The bottom line is you get a single point of access to manage complex chatbot logic without ever leaving your AI client's environment.
Who Is ChatFly MCP For?
This MCP is for the technical founder or operations engineer who knows that customer support needs more than just a basic FAQ page. You need automated, adaptable bots that actually sound human and stay current with product changes.
Needs to instantly update bot knowledge bases after a product launch or policy change, without manually re-writing chat responses.
Requires integrating custom-trained AI assistants into internal tools and monitoring session logs directly from the IDE.
Manages multiple specialized bot workflows, ensuring that lead qualification funnels hit specific criteria before passing off to a human agent.
What Changes When You Connect
- Instantly update bot knowledge bases: If the support team changes a policy, you use
upload_data_sourceto feed the new info, so your bots stay current without human intervention. - Orchestrate multiple agents: Instead of managing one generic chatbot, you can use
list_botsandcreate_botto build specialized instances for billing, sales, or support. - Test interactions on the fly: Use the
chattool to send test questions to a bot and see its live response before deploying it to customers. - Maintain visibility over data: The
list_data_sourcestool lets you audit exactly which documents are feeding knowledge into your system, keeping everything transparent. - Refine bots without downtime: If the welcome message is wrong or needs tweaking, use
update_botto fix it instantly rather than rebuilding the whole bot.
Real-World Use Cases
Handling a product recall announcement
A technical PM needs all bots updated immediately. They instruct their agent to run upload_data_source using the official press release URL, ensuring every bot instantly knows the correct procedure for returns and replacements.
Separating sales from support functions
A founder wants two distinct bots. They use create_bot to make 'Sales Bot' and another to make 'Support Bot'. This allows the agent to route lead qualification questions to one, and technical FAQs to the other.
Monitoring performance after a feature drop
A developer needs to know if the bot is still answering correctly. They use get_bot for details and then run chat with tricky edge-case questions to confirm accuracy.
Onboarding new staff into the bot system
A support manager needs to give a bot access to internal HR documents. They use upload_data_source and specify the document path, allowing the agent to query internal policies via the chatbot.
The Tradeoffs
Treating bots like simple FAQs
Just pasting a static block of text into the bot's welcome message and expecting it to handle complex lead qualification. This fails because the bot has no external knowledge.
→
Use upload_data_source first, pointing at your entire product documentation site. Then use create_bot to give it a specific persona, ensuring its answers are grounded in verifiable data.
Forgetting which bot is which
Trying to modify Bot A's welcome message but accidentally hitting the wrong ID and changing Bot B's settings. This causes confusion across your support channels.
→
Always start by running list_bots to verify all IDs, then use get_bot on the target bot before executing any changes with update_bot.
Training bots in chunks
Only training the bot on one PDF at a time. If you miss an important source, the bot will give contradictory or incomplete answers.
→
Use list_data_sources to keep track of everything. When ready, use upload_data_source with multiple sources (URLs and documents) in one go for comprehensive training.
When It Fits, When It Doesn't
You need this MCP if your problem is 'my chatbot needs to know things that change often' or 'I need more than a simple static FAQ'. You must use it when you need the bot to answer questions based on specific, external knowledge (like product manuals or website content). Don't use it if all you need is a single-purpose chat widget connected to one fixed set of rules—a basic messaging tool might suffice. If your main goal is complex orchestration across many different types of tools (e.g., connecting the chatbot output directly into a CRM record), you should look at dedicated workflow automation platforms, not just bot management.
Common Questions About ChatFly MCP
How do I train ChatFly using `upload_data_source`? +
You give the agent a URL or document path. The bot then crawls or ingests that content to build knowledge. You should run this before testing any new information.
Can I change the welcome message using `update_bot`? +
Yes, you use update_bot and provide the new text. This changes the bot's introduction without affecting its knowledge base or overall functionality.
What if I need to test a bot’s response? Do I use `chat`? +
Yes, running chat lets you simulate a live conversation. It sends your prompt and returns the bot's immediate, accurate response for testing.
How do I see what data my bots are trained on? Do I need `list_data_sources`? +
You use list_data_sources. This command gives you a clear directory of every URL and document that is currently feeding knowledge into your entire chatbot fleet.
How do I use `create_bot` to set up a new chatbot? +
It initializes a brand-new bot instance for you. You must provide both a unique name and an initial welcome message when calling this tool. Remember that creation only sets the shell; you'll need to train it with data sources later.
If I want details about just one specific chatbot, should I use `get_bot`? +
Yes, using get_bot retrieves all metadata for a single specified bot ID. This is helpful when you need to check the current status or configuration of one bot without listing every other instance.
How do I see an overview of all my available bots at once using `list_bots`? +
It returns a directory containing the ID and name of every chatbot you have built. This is the quickest way to confirm how many bot instances are active in your account.
Beyond changing the welcome message, what other settings can I adjust using `update_bot`? +
You can modify various parameters of an existing chatbot instance with this tool. This includes adjusting internal metadata or making operational changes without needing to recreate the entire bot.
How do I find my ChatFly API Key? +
Log in to your account, navigate to Account Settings > API Keys, and generate a new key for your integration.
Can I train a bot on a specific URL? +
Yes! Use the upload_data_source tool and provide the bot ID along with the website URL you want the bot to learn from.
How do I test a bot's response via AI? +
Use the chat tool to send a message to a specific bot ID. Your agent will return the high-fidelity AI response generated by ChatFly.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.