Cody AI MCP. Query private documents with trained bots.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
Cody AI MCP connects your agent directly to specialized internal bots trained on proprietary documents. It lets you manage bot configurations, import web content into knowledge bases, and run complex conversations against company policy guides or HR handbooks.
What your AI agents can do
Create conversation
Starts a new chat session with a specific knowledge base bot.
Get bot details
Retrieves detailed information about one specific bot, showing its history and training data.
Get document status
Checks if a document has finished syncing so the AI can use it for answers.
List every active bot configured in your account so you know what resources are available for querying.
Get specific details about a single bot, confirming which documents it’s trained on and how it's set up.
Pull material from a public URL and add that text directly into the knowledge base for training.
List all folders or individual documents in the knowledge base to audit what information is available.
Verify if a recently uploaded document has finished syncing and is ready for the AI to learn from.
Start a fresh chat session with any specific bot you want your agent to talk to.
Ask AI about this MCP
Supported MCP Clients
OAuth 2.0 Compatible Gemini Waiting for input…
Cody AI with 10 Tools
You can use these tools to list available bots, import web content, check document status, and send messages to your trained AI assistants.
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 Cody AI on Vinkius019d7576create conversation
Starts a new chat session with a specific knowledge base bot.
019d7576get bot details
Retrieves detailed information about one specific bot, showing its history and training data.
019d7576get document status
Checks if a document has finished syncing so the AI can use it for answers.
019d7576import webpage
Adds content from a specific URL into your knowledge base for bot training.
019d7576list bots
Retrieves a list of all available bots configured in your account.
019d7576list conversations
Gets a record of recent chat sessions to help you track history.
019d7576list documents
Retrieves a list of all documents stored in your knowledge base, organized by folder.
019d7576list folders
Retrieve a list of folders in your knowledge base
019d7576list messages
Shows the message history for one specific conversation thread.
019d7576send message
Sends a prompt to the AI, generating an answer based on the bot's knowledge base.
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 Cody AI, then connect any of our 4,800+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 4,800+ 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 Cody AI. 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 10 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
Tracking down answers across siloed documents is a nightmare.
Right now, if an employee asks about a new policy, someone has to manually check the HR intranet, cross-reference the latest wiki guide, and then compare that against the legal department's PDF. It’s copy-pasting between three different tabs, waiting for document versions to sync up.
With this MCP, your agent handles it all automatically. You tell it what topic you care about, and it pulls answers from a dedicated bot trained only on those documents. You get the answer immediately, plus a direct citation showing exactly which file provided the info.
The Cody AI MCP gives your agent access to specialized bots.
You don't just send data; you manage it. You can run `list_bots` first to confirm that a 'Product FAQ' bot exists, then use `import_webpage` if the product team added new guides. This structure means you are building and controlling the knowledge base from the ground up.
This isn't just querying; this is governance. You build reliable intelligence by making sure every piece of data flows through a controlled process.
What you can do with this MCP connector
This connection gives your agent the ability to talk to custom-trained AI assistants built from your own files. Need to know what the latest compliance guide says? Your agent can pull answers from a specific bot trained only on those legal documents, then check where that information came from. You can tell it to build out its knowledge by sending in entire web pages or uploading existing file libraries.
If you're building complex support workflows, this MCP lets your agent manage the whole process—from listing available bots to starting a conversation and tracking every message sent. Because the data is so sensitive, everything runs through Vinkius, which guarantees that credentials pass through a zero-trust proxy; your keys never sit on disk, keeping all corporate information locked down while still giving the AI instant answers.
019d7576-b8d1-7185-b492-786aa5545bc6 
How Cody AI MCP Works
- 1 First, tell the agent which bots are active or what document folders exist.
- 2 Next, send content—either a web page via
import_webpageor asking for the status of uploaded files usingget_document_status—to expand the bot's knowledge. - 3 Finally, start a conversation with the desired bot and use
send_messageto get answers based on the entire corpus.
The bottom line is you manage your proprietary data source through structured commands so your AI can answer questions accurately and cite its sources.
Who Is Cody AI MCP For?
The technical knowledge manager who dreads manually updating dozens of internal wikis. The support engineer tired of copy-pasting answers from old PDFs. Developers needing to embed a private, accurate Q&A system into an application.
Using the agent to query trained bots for instant answers on customer issues, pulling data directly from policy guides.
Managing bot training by importing new web content and listing documents to ensure the knowledge base is current.
Building automated processes that first check a bot's status, then list conversations, and finally send messages in sequence.
What Changes When You Connect
- Stop guessing where to find answers. You can use
list_botsandget_bot_detailsto map out exactly which bot owns which set of company policies, making sure the right information is queried every time. - Keep your knowledge base current without manual data entry. Use
import_webpageto pull fresh content from a URL instantly, keeping your bots trained on live web material. - Audit everything that happens in the support center. You can use
list_conversationsand thenlist_messagesto track message history across all user interactions for compliance checks. - Don't wait for documents to finish uploading. Use
get_document_statusbefore querying so your agent only tries to answer questions once the content is fully indexed and ready. - Build complex automation chains that manage multiple pieces of data: start by listing folders, then sending a message using
send_message, all through one cohesive flow.
Real-World Use Cases
The Support Agent needs an immediate answer on PTO policy.
Instead of searching multiple shared drives, the agent uses its AI client to query a specific 'HR Bot' via send_message. The bot immediately replies with the correct policy details and cites the exact document source.
The Compliance team adds a new regulation.
A knowledge manager doesn't manually update every guide. They use the agent to run import_webpage on the new regulation URL, expand the 'Legal Bot,' and wait for confirmation via get_document_status before it goes live.
The Developer needs an audit trail of recent user queries.
A developer uses the agent to run list_conversations, then checks the message history using list_messages. This lets them build a dashboard that tracks what users are asking about most often.
The Product Team needs to see how many internal bots exist.
They run list_bots immediately. This gives them an overview of all specialized AI assistants, confirming whether they need a new bot or just more documents for the existing ones.
The Tradeoffs
Asking the AI to read everything at once.
Your agent tries to answer a question using every single document in the knowledge base, leading to vague answers and confusion about source material.
→
First, use list_bots to identify the one bot trained specifically on that topic. Then, send the message via send_message targeted only at that specific bot.
Assuming content is ready immediately.
A user uploads a massive PDF and immediately asks the agent a question, resulting in an error because the AI hasn't indexed the new data yet.
→
After uploading any file or webpage, always use get_document_status to confirm that the content has finished syncing before triggering a conversation.
Forgetting the context of the chat.
The agent pulls an answer but can't show where it got the info, making the data useless for auditing or compliance purposes.
→
Always check list_messages after a conversation to verify that the response included source document citations and file names.
When It Fits, When It Doesn't
Use this MCP if your core problem is answering complex questions based on private, proprietary documents (like internal wikis or legal guides). You need control over which data set answers the question. Don't use it if you just need a general-purpose chatbot that pulls from the public web; those generic tools won't know your company's policies. Use this MCP when you need to manage the lifecycle of knowledge: importing content, checking its readiness, and finally querying it via specialized bots.
Common Questions About Cody AI MCP
How do I know if my new documents are ready to be used by the bot? +
Use get_document_status. This tool checks the syncing status, telling you exactly when the AI has finished learning the content and it's safe to ask questions.
What is the difference between listing documents and listing messages with Cody AI? +
Listing documents (list_documents) shows files in your knowledge base. Listing messages (list_messages) shows the text history of a specific chat conversation.
Can I make multiple bots talk to each other using this MCP? +
While the tools don't chain conversations directly, you can build workflows that first use list_bots to identify the needed source, and then send messages sequentially to different specialized bots.
How do I get a list of all available chatbots with Cody AI? +
Run list_bots. This retrieves every active bot configured in your account, giving you an immediate overview of your current knowledge resources.
How do I check my knowledge base structure before running queries with `list_folders`? +
Use list_folders to view your entire organizational hierarchy. This shows you all the distinct categories where documents live, helping you target specific content groups for better query results and context setting.
What happens if I use `send_message` repeatedly or exceed usage limits? +
The system handles API rate limiting. If your agent hits a ceiling, it will return an error indicating the limit was reached. You'll need to build backoff logic into your automation workflow to pause and retry later.
What specific information can I pull about a single bot using `get_bot_details`? +
This tool pulls detailed metadata for one bot. You get its configuration, the total number of documents it references, and key settings like when it was last updated. This is vital for auditing or troubleshooting.
How does `import_webpage` handle large amounts of content from a single URL? +
The tool sends the entire web page's text to be vectorized and trained by Cody AI. Processing time depends on the sheer length of the content, so always check the document status afterward to confirm training is complete.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.