Provides built-in kubectl port-forward management with auto-restart capabilities for connecting to Unblu deployments running in Kubernetes clusters.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Unblu MCP Serverlist all conversations from yesterday"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
unblu-mcp
A Model Context Protocol (MCP) server for debugging and operating Unblu deployments. Optimised for debugging workflows — curated typed tools for common operations, plus an escape hatch for the full 300+ endpoint API.
📚
Design
The server exposes three layers, each progressively more powerful:
Layer | What it is | When to use |
Curated tools | Typed, token-efficient tools for common debugging tasks | 90% of debugging sessions |
| Generic escape hatch for any of 331 Unblu API operations | When a curated tool doesn't exist |
Resources | Read-only | Discovery and schema inspection |
Quick Start
Installation
Or install persistently: uv tool install unblu-mcp
MCP Client Configuration
Direct API access with an API key:
Kubernetes port-forward (auto-managed):
Tools
Curated tools (read-only)
Tool | Description |
| Current account info — good first call to verify connectivity |
| Search conversations with filters |
| Full conversation detail with participants |
| Find persons (visitors, agents) |
| Full person detail |
| Batch-resolve up to 20 persons in parallel |
| Find registered users |
| Full user detail |
| Check agent availability per named area |
| List named areas (routing targets) |
| Discover API operations by keyword |
Mutation tools
Tool | Description |
| Assign a conversation to an agent |
| End an active conversation |
Escape hatch
Tool | Description |
| Execute any of the 331 Unblu API operations |
Resources
URI | Description |
| JSON list of all API service groups |
| Full resolved schema for a specific operation |
Prompts
Prompt | Description |
| Step-by-step debugging workflow for a conversation |
| Locate an agent and check their availability |
| Validate account configuration and connectivity |
Safety
By default, execute_operation blocks DELETE operations unless confirm_destructive=True is passed. For stricter control, install the optional Eunomia policy engine:
Then point to a policy file:
The bundled config/mcp_policies.json allows all read-only and curated tools while blocking destructive execute_operation calls by operation ID pattern.
Development
License
ISC License