> ## Documentation Index
> Fetch the complete documentation index at: https://docs.bindo.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Connect your agent to Bindo

> Read enquiries, maintain knowledge and prepare replies through MCP.

Bindo serves Streamable HTTP MCP at your workspace API origin followed by `/mcp`. Find the exact URL in **Settings → Agent access**. The intended hosted origin is `https://app.bindo.ai`; use the address supplied during onboarding.

For a local development instance:

```toml theme={null}
[mcp_servers.bindo]
url = "http://127.0.0.1:8080/mcp"
bearer_token_env_var = "BINDO_MCP_TOKEN"
```

Set `BINDO_MCP_TOKEN` privately in the client environment. Tools are filtered by token permissions. Read the `bindo://guide` resource when connecting.

### A useful workflow

1. Call `list_enquiries` with `queue: "unanswered"`. Follow `next` even after an empty filtered page.
2. Call `get_enquiry_context` for the selected enquiry. Check blockers and the latest inbound.
3. Inspect knowledge and assistant settings. Call `generate_reply_draft`, or prepare a manual draft with `create_reply_draft`.
4. Return the saved draft ID and review reasons to staff. Do not claim the reply was sent.

Configuration tools include `save_knowledge`, `configure_assistant`, `configure_reply_rule`, email setup, `configure_integration` and `discover_integration_tools` when explicitly delegated.

There is no MCP tool for sending a reply, granting connected-tool permissions or purchasing a subscription. Those are human decisions.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.