# Cognita MCP for AI Agents AI Agent Connect > Cognita (RAG Framework) lets you manage your modular RAG pipelines through natural conversation with your AI agent. You can sync data from remote sources, audit your vector store chunks, and query your knowledge base directly. It's built for people who need to manage complex RAG workflows without jumping between different dashboards or writing custom scripts. ## Overview - **Category:** friends-mcp - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_SHxG8KJRFQw2m3lX2GfeLS7zRF34WrsRFB0j95Nv/ai-agent-connect - **Tags:** rag-framework, vector-search, embedding-models, data-ingestion, ai-pipeline, knowledge-retrieval ## Description Cognita lets you manage your modular RAG workflows through natural conversation with your AI agent. It gives you a way to handle the backend of your knowledge base without jumping between different dashboards or writing custom scripts. You can sync data from remote sources, audit your vector store chunks, and query your knowledge base directly. It makes the backend of your AI applications feel like a conversation rather than a series of manual configuration steps. Instead of manually checking if a file synced or trying to figure out why a specific chunk was retrieved, you can just ask your agent to do the heavy lifting. You can force a sync from a remote API, audit the specific text segments your agent is seeing, or list every model registered in your installation to ensure your endpoints are live. This means you don't have to worry about the underlying plumbing of your vector space while you're trying to ship a product. It handles the heavy lifting of ensuring your knowledge base stays updated and your retrieval logic stays sharp. If you're looking for a way to manage complex RAG pipelines without the usual headache of manual oversight, this is the way to do it. You can find this and thousands of other tools in the Vinkius catalog to build out your full AI stack. ## Tools ### list_data_sources Use this to see all the external buckets and APIs mapped to your AI workflows. It confirms your data pipelines are connected. ### ingest_data Use this to trigger a sync to pull new files into your vector space. It handles the heavy lifting of generating new resource directories. ### rag_query Use this to ask a question and get a synthesized answer from your vector store. It pulls relevant context to give you a grounded response. ### search_chunks Use this to perform a search to find specific text segments in your vector store. It lets you verify exactly what your agent knows. ### list_models Use this to see every LLM and embedding model registered in your Cognita instance. It helps you verify which models are ready for production. ### list_collections Use this to see all your RAG collections and their specific configurations. It helps you keep track of different knowledge domains. ### get_collection Use this to pull specific logging and payload IDs for a single collection. It is great for deep-diving into specific data sets. ## Prompt Examples **Prompt:** ``` List all RAG collections in Cognita ``` **Response:** ``` I found 3 collections: 'technical-docs', 'legal-kb', and 'customer-support'. Which one would you like to inspect for metadata? ``` **Prompt:** ``` Query collection 'technical-docs' for: 'How do I configure OAuth in our API?' ``` **Response:** ``` Based on your technical docs, you need to navigate to the /auth/settings endpoint and register a new client ID. [Detailed answer synthesized from 3 context chunks]. ``` **Prompt:** ``` Ingest data from source 'gh-repo-vinkius' into collection 'technical-docs' ``` **Response:** ``` Ingestion pipeline triggered! Cognita is now syncing 'gh-repo-vinkius' into the 'technical-docs' collection. I will let you know once the knowledge base is updated. ``` ## Capabilities ### Sync remote data sources Pull files from SQL, Cloud Storage, or APIs into your vector space automatically. ### Audit RAG collections Inspect embedding configurations and token lengths across your knowledge base. ### Query vector stores Get synthesized answers from your stored context using natural language questions. ### Verify document chunks Perform lexical or semantic searches to pull and inspect raw text segments. ### Monitor model registries See every LLM and embedding model registered in your Cognita instance. ### Manage data sources List all connected external data sources mapped to your AI workflows. ## Use Cases ### Verifying chunking logic A data scientist needs to check if a new PDF was correctly chunked. They ask the agent to search for specific text segments using `search_chunks` to verify the results. ### Checking model availability An AI engineer wants to see which models are currently live in production. They ask the agent to list all registered models with `list_models` to confirm the setup. ### Auditing data sources A product manager wants to know what data is being fed into the Technical Docs collection. They use `list_data_sources` to see all connected APIs and buckets. ### Manual data ingestion A developer needs to quickly test if a new SQL source is correctly syncing. They use `ingest_data` to trigger a manual sync and then check the collection. ## Benefits - Keep your knowledge base fresh by using `ingest_data` to sync SQL and Cloud Storage files automatically. - Verify your retrieval accuracy by using `search_chunks` to pull and inspect raw text segments in your vector store. - Audit your RAG collections with `list_collections` to ensure your embedding configurations and token lengths are correct. - Get synthesized answers from your vector store using `rag_query` to test your RAG pipeline in real-time. - Monitor your active models with `list_models` to confirm that your LLM endpoints are reachable and active. - Manage your entire data pipeline by using `list_data_sources` to see every external mapping in one place. ## How It Works The bottom line is you get a direct command line for your RAG infrastructure inside your favorite AI chat interface. 1. Subscribe to this Connector and grab your Cognita Base URL and API Key. 2. Plug those credentials into your AI client. 3. Start asking your agent to sync files, audit chunks, or query your knowledge base. ## Frequently Asked Questions **Can the Cognita MCP sync my SQL data automatically?** Yes, it can. The Connector allows your agent to trigger data ingestion from various sources, including SQL databases, cloud storage, and remote APIs, to keep your knowledge base updated. **How do I check if my documents are being chunked correctly with Cognita?** You can use the search tool to pull raw document chunks from your vector store. This lets you see the exact text segments your agent uses to answer questions. **Can I see which LLMs are active in my Cognita instance?** Yes, you can ask your agent to list the models. This shows you every LLM and embedding model currently registered in your Cognita setup. **Does Cognita work with my existing data sources?** It works with many common sources like Cloud Storage, SQL, and APIs. You can list your connected sources to verify which ones are mapped into your workflows. **How does Cognita help with RAG debugging?** It lets you audit your collections and search specific chunks. By seeing the raw data and the model registry in one place, you can quickly find where a retrieval pipeline is failing. **Can I query my knowledge base directly through the Cognita MCP?** Yes, you can. Your agent can perform RAG queries that search your vector store and synthesize accurate answers based on your stored context. **Can my agent perform semantic RAG queries against my collections?** Yes. The 'rag_query' tool allows you to ask questions in natural language. The agent queries your vector store via Cognita and uses an LLM to synthesize a final answer based explicitly on the retrieved context. **How can I trigger a data ingestion pipeline through the agent?** Provide the collection name and the data source FQN (Fully Qualified Name). The 'ingest_data' tool will command the Cognita backend to start a sync, updating your RAG vector space with the latest remote documents. **Can I audit the raw document chunks before LLM generation?** Absolutely. Use the 'search_chunks' tool to perform vector searches that return raw text segments and metadata without LLM synthesis. This is the perfect way to verify that your retrieval logic is pulling the correct data boundaries.