← Back to all guides

MCP Connectors & the Connect Page — Wire AI Agents to 3,000+ Apps

Langoedge Team9 min read

What is an MCP Connector?

An MCP Connector is a node you drop on a Text Graph that hands your agent every tool a third-party app exposes — Slack, Gmail, Airtable, HubSpot, Notion, Stripe, and 3,000+ more — without anyone writing an integration for it first.

Definition — MCP (Model Context Protocol): An open standard for describing the tools an application offers, so an AI agent can discover and call them at runtime instead of relying on hand-written wrappers.

Langoedge resolves connectors against a hosted MCP catalogue covering the full third-party app directory. Two consequences worth understanding up front:

  • The catalogue is discovered live, not hardcoded. When an app adds an action, or a new app joins the directory, it shows up in your picker without a Langoedge release.
  • Adding an app is configuration, not engineering. Pick the app, pick the tools, attach the connector to a step. There is no code and no waiting for a connector to be built.

Two steps, always in this order

1. **Connect the account** on the **Connect** page — this is the OAuth handshake that gives Langoedge permission to act as you in that app. 2. **Add an MCP Connector node** to your graph and attach it to a step — this is what actually gives the agent the tools.

Skipping step 1 doesn't break the builder; it just means the connector can't list any tools yet.


Part 1 — The Connect Page

Go to Connect in the dashboard. This page is the credential vault for your whole workspace: every account you link here is scoped to your user and reusable across every graph you own.

Linking an account

1

Find the app

Start typing in the **App to connect** field. The list is the live catalogue, so it is far too large to render at once — only the first 100 matches are shown, and typing narrows the search on the server. If you can't see an app, keep typing rather than scrolling.
2

Click Connect Account

A secure connect flow opens in an overlay. You authenticate directly with the provider — Google, Slack, whoever it is. Langoedge never sees your password.
3

Confirm it landed

The account appears as a card below the picker, showing the app icon and the account name. That card is your proof the link is live.

Disconnecting

Each account card has a Disconnect button. Disconnecting revokes Langoedge's access immediately and clears the cached tool catalogue, so any graph step relying on that app will start reporting a missing connection on its next run.

One account per app, per user

Accounts are scoped to your Langoedge user. Two people on the same workspace each connect their own Slack. A graph runs against its owner's connected accounts, not the accounts of whoever triggers it — which matters when you publish a graph or embed a widget: the actions are performed as the graph owner.


Part 2 — Required Connections

Some connections aren't optional extras — the platform gates on them.

**OpenAI is required to use Langoedge at all.** If no OpenAI account is connected, the dashboard shows a non-dismissable overlay on every page until you connect one. This is the single most common reason a new account appears "frozen" — it isn't, it's waiting on the Connect page.

Beyond that, the graph editor checks the models your graph actually uses and blocks the canvas if the matching provider isn't connected. The mapping is by model name prefix:

Model name looks like Provider you must connect
gpt-…, o1…, o3…, o4… OpenAI
claude…, anthropic/… Anthropic
gemini-…, gemini/…, google/… Google Gemini
llama-…, mixtral-…, gemma-…, groq/…, meta-llama/… Groq
Anything ending -us or -aus None — these are Azure-hosted and need no connection

The Knowledge Store additionally requires OpenAI, since it uses OpenAI embeddings to index your files.


Part 3 — Adding an MCP Connector to a Graph

Open a Text Graph and add an MCP Connector node to the canvas.

An MCP Connector is a **tool-definition node**, like the Retriever and API Tool nodes. It has no input or output handles and is never wired with edges. It sits on the canvas defining a capability, and a step opts into it from its own tool list.
1

Pick the app

Open the node and choose an app from the **App** dropdown. Selecting one immediately loads that app's tool catalogue.
2

Choose which tools to expose

By default the connector exposes **every tool the app offers** — this is the common case and costs nothing extra. Use the **Enabled tools** dropdown to turn individual tools off if you want to keep an agent narrowly scoped, for example exposing Slack's read tools but not its delete tools. The node card summarises the result as *"Using every tool this app offers"* or *"Using all tools except …"*.
3

Attach it to a step

Open the step that should be able to act on the app, and select the connector from that step's tool list. A step can use several connectors at once, and one connector can be attached to several steps.

If the app isn't connected yet

If you haven't linked the account, the node shows "No connected account yet" with a direct Connect link that opens the right connect flow. Follow it, then press Load Tools again.

Refreshing a catalogue

Tool catalogues are cached for an hour. If a provider ships a new action and you want it now, press Refresh on the node to force a fresh fetch.


How Tool Resolution Works

Understanding the read path explains most of the behaviour you'll see:

flowchart LR A[Graph opens picker / compiles] --> B{Redis cache} B -- hit and fresh --> Z[Tools] B -- miss or stale --> C{MongoDB} C -- hit and fresh --> Z C -- miss or stale --> D[Upstream MCP catalogue] D -- success --> E[Write through Mongo + Redis] --> Z D -- failure --> F[Serve last known good copy] --> Z

Three things follow from this:

  • Opening the tool picker is fast. It's a cache read, not an upstream round trip.
  • An upstream outage degrades rather than breaks. If the upstream fetch fails, Langoedge serves the last catalogue it successfully fetched, however old. A slightly stale schema is recoverable; a graph that compiles with zero tools just looks like a broken agent.
  • A catalogue is shared, not per-graph. Tool names, descriptions, and schemas are properties of the app, not choices you made, so one entry is shared by every graph and every user. Only your decisions — which app, which tools withheld — are stored on your graph.

If an app's tools genuinely cannot be resolved from any layer, that connector is skipped and the step runs without it, with a warning in the logs rather than a hard failure.


What Is Not an MCP Connector

A handful of integrations are deliberately kept as hand-written modules rather than going through MCP, because they need behaviour beyond a generic tool call — retries, custom formatting, domain-specific aggregation, or a real-time voice path:

Integration Why it's hand-written
Cliniko Healthcare scheduling with domain-specific aggregation across practitioners, appointment types, and availability.
ServiceM8 Field-service job and client workflows.
Twilio / Telnyx Telephony and SMS sit on the real-time voice path, not the tool-calling path.
Pinecone / Qdrant Vector stores backing the Knowledge Base and Retriever tools.

Native tools — the web crawler, web search, academic search, the SQL toolkit, the sandboxed code runner, and RAG retrievers — are unaffected by any of this and need no account connection beyond their own configuration.

**Voice Graphs do not support MCP connectors.** The connector node exists only on Text Graphs. A voice agent reaches an MCP-backed app by calling a Text Graph as a tool — which is the recommended pattern anyway, since it keeps slow third-party calls off the live audio loop. See the [Voice Graph guide](/guides/voice-graph) for how to mount a Text Graph inside a voice session.

Security

  • Credentials are held by the connect layer and referenced by an opaque account id. Langoedge stores the link, not your password.
  • The LLM never sees your credentials. Authentication is attached server-side, per request, at the moment a tool is invoked. The model sees only the tool's inputs and whatever the app returns.
  • Tokens are minted fresh for each call rather than baked into a compiled graph, because graphs are compiled once and cached but invoked arbitrarily later.
  • Every call is scoped to the graph owner's connected accounts. A graph cannot reach another user's linked apps.

Troubleshooting

The whole dashboard is blurred behind a 'Connection Required' overlay
You have no OpenAI account connected. Go to **Connect**, link OpenAI, and the overlay clears. The same overlay appears on a specific graph if that graph uses a model whose provider — Anthropic, Google Gemini, or Groq — isn't connected.
The app I want isn't in the dropdown
Only the first 100 matches render at a time, because the full catalogue runs to 3,000+ apps and mounting all of them freezes the picker. Keep typing — the search runs server-side across the whole directory.
The connector says 'No connected account yet'
Langoedge can see the app but you haven't authorised it. Use the **Connect** link shown on the node, or go to the Connect page and link the account, then press **Load Tools** again.
A tool the app definitely has isn't listed
The catalogue is cached for an hour. Press **Refresh** on the connector node to force a fresh fetch.
My agent stopped being able to use an app
Check the Connect page — if the account was disconnected or its authorisation was revoked at the provider, the tools stop resolving. Reconnect and the connector recovers on its next fetch.
Do I need Zapier or Make?
No. Earlier documentation suggested routing through automation platforms via webhooks to reach apps Langoedge didn't natively support. That was true before MCP connectors landed. Today the connector reaches the whole catalogue directly, and API Webhooks are for genuinely custom REST endpoints rather than for reaching mainstream SaaS apps.
Can I use MCP tools in a Voice Graph?
Not directly. Attach the connector to a Text Graph, then mount that Text Graph as a tool inside your Voice Graph node. This also keeps a slow third-party call from stalling the conversation, since the voice agent can run the Text Graph asynchronously and keep talking.

LT

Langoedge Team

The Langoedge engineering team builds AI agent infrastructure that empowers businesses to deploy reliable, observable AI staff. Follow Langoedge Team on LinkedIn for product updates and architectural deep dives.