Unifi MCP Tool Builder

UniFi MCP Tool Builder

Overview

This skill guides you through adding new MCP tools to the existing UniFi MCP Server. Unlike building an MCP server from scratch, this focuses on extending the current 74-tool implementation with new UniFi Network Controller functionality while maintaining the project's quality standards.

Project Context:

• Existing Tools: 74 MCP tools across 9 feature categories
• Test Coverage: 990 tests, 78.18% overall coverage
• Architecture: FastMCP + Pydantic v2 + async/await patterns
• API Support: Local gateway (full), Cloud V1/EA (limited)
• Quality Standards: 80% minimum test coverage, TDD required

Process

Phase 1: Research and Planning

1.1 Understand the Tool Request

Before implementation, clarify:

• What UniFi feature/endpoint needs to be exposed?
• Which API mode supports it (local, cloud-v1, cloud-ea)?
• What workflow does this enable for AI agents?
• How does this fit into existing tool categories?

Tool Categories in UniFi MCP Server:

1. Device Management (list, control, upgrade)
2. Client Management (block, unblock, statistics)
3. Network Configuration (VLANs, subnets, DHCP)
4. Firewall & Security (rules, zones, ACLs)
5. WiFi/SSID Management (create, configure, statistics)
6. QoS & Traffic (profiles, flows, DPI)
7. Backup & Restore (automated backups, restore)
8. Site Management (multi-site, aggregation)
9. Topology & Monitoring (network maps, health)

1.2 Study UniFi API Documentation

Load UniFi API reference:

# Read the comprehensive UniFi API documentation
Read: docs/UNIFI_API.md

This document contains verified endpoints for UniFi Network v10.0.156+.

Critical considerations:

• Endpoint availability: Not all documented endpoints exist in all versions
• API mode differences: Local gateway vs Cloud APIs have different capabilities
• Response formats: Local API returns different structures than Cloud API
• Authentication: Local uses local credentials, Cloud uses API keys

1.3 Review Existing Patterns

Study similar tools in the codebase:

# Example: For a new device tool, read existing device tools
Read: src/tools/devices.py
Read: src/tools/device_control.py

# For network tools:
Read: src/tools/networks.py
Read: src/tools/network_config.py

# For testing patterns:
Read: tests/unit/test_devices.py

Key patterns to identify:

• How similar tools structure their inputs (Pydantic models)
• Response formatting (JSON with TypedDict or Pydantic response models)
• Error handling patterns
• Confirmation requirements for mutating operations
• Dry-run mode implementation
• Caching strategies (if applicable)

1.4 Verify API Endpoint Availability

CRITICAL: Not all documented API endpoints actually exist!

The UniFi MCP Server has discovered that many documented endpoints don't exist in real controllers. Before implementation:

1. Check endpoint existence in docs/UNIFI_API.md verification notes
2. Test on real hardware if possible (U7 Express, UDM Pro, etc.)
3. Document findings in code comments and API.md

Known endpoint issues:

• ZBF matrix endpoints don't exist (use Firewall Policies v2 instead)
• Some statistics endpoints are documented but not available
• Cloud APIs have very limited endpoint support vs local gateway

1.5 Create Implementation Plan

Document your plan covering:

Tool Definition:

• Tool name (follow naming convention: {action}_{resource})
• Description (one-line summary for LLMs)
• Input parameters (with Pydantic model)
• Response format (TypedDict or Pydantic model)
• Error scenarios

Data Models:

• Pydantic models for request validation
• Response models (if complex data structure)
• Enums for allowed values

Testing Strategy:

• Unit test scenarios (minimum 10-15 tests)
• Mock API responses
• Edge cases (empty results, errors, pagination)
• Target coverage: 85%+ for new code

Documentation:

• Docstring with examples
• API.md section update
• README.md feature list update (if major feature)

Phase 2: Implementation

2.1 Create Pydantic Models (if needed)

Location: src/models/{feature}.py

from enum import Enum
from typing import Optional
from pydantic import BaseModel, Field, ConfigDict

class YourRequestModel(BaseModel):
    """Request model for your_tool operation.

    Attributes:
        site_id: UniFi site identifier
        resource_id: Resource to operate on
        options: Optional configuration parameters
    """
    model_config = ConfigDict(extra="forbid", str_strip_whitespace=True)

    site_id: str = Field(..., description="UniFi site ID (e.g., 'default')")
    resource_id: str = Field(..., min_length=1, description="Resource identifier")
    options: Optional[dict] = Field(None, description="Additional options")

class YourResponseModel(BaseModel):
    """Response from your_tool operation."""
    success: bool
    resource_id: str
    message: str

Key requirements:

• Use Pydantic v2 syntax (model_config instead of Config)
• extra="forbid" to reject unknown fields
• Comprehensive Field descriptions for LLM understanding
• Type hints for all fields
• Docstrings explaining purpose

2.2 Implement the MCP Tool

Location: src/tools/{category}.py (or new file if new category)

from fastmcp import Context
from src.models.your_model import YourRequestModel, YourResponseModel
from src.api.client import UniFiClient
from src.utils.validators import validate_site_id

@mcp.tool()
async def your_tool_name(
    site_id: str,
    resource_id: str,
    options: dict | None = None,
    settings: Context = None,
    confirm: bool = False,
    dry_run: bool = False,
) -> dict:
    """Brief one-line description of what this tool does.

    Detailed explanation of the tool's purpose, when to use it,
    and what it accomplishes. Include examples of common use cases.

    Args:
        site_id: UniFi site identifier (e.g., 'default')
        resource_id: Resource to operate on
        options: Optional configuration dictionary
        settings: FastMCP context (auto-injected)
        confirm: Required for mutating operations (default: False)
        dry_run: Preview changes without applying (default: False)

    Returns:
        Dictionary containing:
        - success: Operation success status
        - resource_id: Modified resource identifier
        - message: Human-readable result message
        - details: (if dry_run) Preview of changes

    Raises:
        ValueError: If site_id is invalid or resource not found
        PermissionError: If confirm=True not provided for mutation

    Examples:
        >>> # Read-only operation
        >>> result = await your_tool_name(site_id="default", resource_id="abc123")

        >>> # Mutating operation (requires confirmation)
        >>> result = await your_tool_name(
        ...     site_id="default",
        ...     resource_id="abc123",
        ...     confirm=True
        ... )

        >>> # Preview changes first
        >>> preview = await your_tool_name(
        ...     site_id="default",
        ...     resource_id="abc123",
        ...     dry_run=True
        ... )
    """
    # Validate inputs
    validate_site_id(site_id)

    # Get client from context
    client: UniFiClient = settings.get("client")

    # MUTATING OPERATIONS: Require confirmation
    if not dry_run:
        if not confirm:
            raise PermissionError(
                "This operation modifies network configuration. "
                "Set confirm=True to proceed, or use dry_run=True to preview changes."
            )

    # Dry-run mode: Preview changes
    if dry_run:
        return {
            "success": True,
            "dry_run": True,
            "preview": {
                "operation": "your_operation",
                "resource_id": resource_id,
                "changes": {"key": "value"},
            },
            "message": "Dry-run completed. Set confirm=True to apply changes."
        }

    # Execute operation
    try:
        response = await client.request(
            method="POST",
            endpoint=f"/api/s/{site_id}/rest/your_endpoint",
            json={"resource_id": resource_id, **(options or {})},
        )

        return {
            "success": True,
            "resource_id": resource_id,
            "message": f"Successfully processed resource {resource_id}",
            "data": response.get("data", {}),
        }

    except Exception as e:
        # Provide actionable error messages
        if "not found" in str(e).lower():
            raise ValueError(
                f"Resource '{resource_id}' not found in site '{site_id}'. "
                f"Use list_resources tool to see available resources."
            ) from e
        raise

Implementation checklist:

• FastMCP @mcp.tool() decorator
• Comprehensive docstring with examples
• Input validation (use existing validators in src/utils/validators.py)
• Confirm requirement for mutating operations
• Dry-run mode support for previewing changes
• Actionable error messages that guide LLM to correct usage
• Async/await for all I/O operations
• Proper exception handling with context

2.3 Register Tool in Main Server

Location: src/main.py

Add your tool to the imports and ensure it's registered:

from src.tools.your_category import your_tool_name

# Tools are auto-registered via @mcp.tool() decorator
# No manual registration needed with FastMCP

2.4 Follow Code Quality Standards

Run quality checks:

# Format code
black src/tools/your_file.py
isort src/tools/your_file.py

# Lint
ruff check src/tools/your_file.py --fix

# Type check
mypy src/tools/your_file.py

Phase 3: Test-Driven Development

3.1 Write Tests BEFORE Implementation

Location: tests/unit/test_{category}.py

CRITICAL: UniFi MCP Server requires Test-Driven Development (TDD)

1. Write failing tests first
2. Implement just enough to make tests pass
3. Refactor and improve
4. Achieve 85%+ coverage for new code

3.2 Test Structure

import pytest
from unittest.mock import AsyncMock, MagicMock, patch
from src.tools.your_category import your_tool_name

class TestYourToolName:
    """Test suite for your_tool_name."""

    @pytest.fixture
    def mock_context(self):
        """Mock FastMCP context with UniFi client."""
        context = MagicMock()
        client = AsyncMock()
        context.get.return_value = client
        return context, client

    @pytest.mark.asyncio
    async def test_successful_operation(self, mock_context):
        """Test successful tool execution."""
        context, client = mock_context
        client.request.return_value = {
            "data": {"_id": "abc123", "name": "test"}
        }

        result = await your_tool_name(
            site_id="default",
            resource_id="abc123",
            settings=context,
            confirm=True
        )

        assert result["success"] is True
        assert result["resource_id"] == "abc123"
        client.request.assert_called_once()

    @pytest.mark.asyncio
    async def test_requires_confirmation(self, mock_context):
        """Test that mutating operations require confirm=True."""
        context, client = mock_context

        with pytest.raises(PermissionError, match="Set confirm=True"):
            await your_tool_name(
                site_id="default",
                resource_id="abc123",
                settings=context,
                confirm=False
            )

    @pytest.mark.asyncio
    async def test_dry_run_mode(self, mock_context):
        """Test dry-run preview mode."""
        context, client = mock_context

        result = await your_tool_name(
            site_id="default",
            resource_id="abc123",
            settings=context,
            dry_run=True
        )

        assert result["dry_run"] is True
        assert "preview" in result
        client.request.assert_not_called()

    @pytest.mark.asyncio
    async def test_invalid_site_id(self, mock_context):
        """Test validation of site_id."""
        context, client = mock_context

        with pytest.raises(ValueError, match="Invalid site"):
            await your_tool_name(
                site_id="",
                resource_id="abc123",
                settings=context
            )

    @pytest.mark.asyncio
    async def test_resource_not_found(self, mock_context):
        """Test handling of not found errors."""
        context, client = mock_context
        client.request.side_effect = Exception("Resource not found")

        with pytest.raises(ValueError, match="not found"):
            await your_tool_name(
                site_id="default",
                resource_id="nonexistent",
                settings=context,
                confirm=True
            )

    @pytest.mark.asyncio
    async def test_with_optional_parameters(self, mock_context):
        """Test tool with optional parameters."""
        context, client = mock_context
        client.request.return_value = {"data": {"_id": "abc123"}}

        result = await your_tool_name(
            site_id="default",
            resource_id="abc123",
            options={"key": "value"},
            settings=context,
            confirm=True
        )

        assert result["success"] is True
        # Verify options were passed to API
        call_args = client.request.call_args
        assert call_args[1]["json"]["key"] == "value"

Test coverage requirements:

• Successful operation (happy path)
• Confirmation requirement (for mutating ops)
• Dry-run mode
• Input validation (invalid site_id, etc.)
• Error handling (not found, API errors)
• Optional parameters
• Edge cases (empty results, pagination)
• Response format validation

Target: 85%+ coverage for new code

3.3 Run Tests

# Run your specific test file
pytest tests/unit/test_your_category.py -v

# Run with coverage
pytest tests/unit/test_your_category.py --cov=src/tools/your_category --cov-report=term-missing

# Ensure coverage meets target (85%+)
pytest tests/unit/test_your_category.py --cov=src/tools/your_category --cov-report=html
open htmlcov/index.html  # Review coverage report

Phase 4: Documentation

4.1 Update API.md

Location: API.md

Add your tool to the appropriate section with complete documentation:

### your_tool_name

**Description**: Brief one-line description

**Purpose**: Detailed explanation of when and why to use this tool

**Parameters**:
- `site_id` (string, required): UniFi site identifier (e.g., "default")
- `resource_id` (string, required): Resource identifier to operate on
- `options` (object, optional): Additional configuration options
- `confirm` (boolean, optional): Required for mutating operations (default: false)
- `dry_run` (boolean, optional): Preview changes without applying (default: false)

**Returns**:
```json
{
  "success": true,
  "resource_id": "abc123",
  "message": "Successfully processed resource",
  "data": { /* resource details */ }
}

Dry-run response:

{
  "success": true,
  "dry_run": true,
  "preview": {
    "operation": "your_operation",
    "resource_id": "abc123",
    "changes": { /* preview of changes */ }
  },
  "message": "Dry-run completed. Set confirm=True to apply changes."
}

Examples:

# Example 1: Preview changes (dry-run)
result = await mcp.call_tool("your_tool_name", {
    "site_id": "default",
    "resource_id": "abc123",
    "dry_run": True
})

# Example 2: Execute operation
result = await mcp.call_tool("your_tool_name", {
    "site_id": "default",
    "resource_id": "abc123",
    "confirm": True
})

Error Handling:

• ValueError: Invalid site_id or resource not found
• PermissionError: Mutating operation requires confirm=True
• APIError: UniFi API returned error (check message for details)

API Endpoint: POST /api/s/{site}/rest/your_endpoint

Supported API Modes:

• ✅ Local Gateway API (full support)
• ⚠️ Cloud V1 API (limited support)
• ❌ Cloud EA API (not supported)

### 4.2 Update README.md (if major feature)

If adding a new feature category, update README.md:

```markdown
### Your New Feature Category

- **Feature Name**: Description of what it enables
- **Tools**: List of new tools
- **Use Cases**: Common scenarios where this is valuable

4.3 Update CHANGELOG.md

Add entry under "Unreleased" section:

## [Unreleased]

### Added
- New tool `your_tool_name` for managing XYZ (#123)
- Support for ABC feature in UniFi Network 10.0+

Phase 5: Quality Verification

5.1 Run Full Test Suite

# Run all tests
pytest tests/unit/

# Verify no regressions
pytest tests/unit/ --cov=src --cov-report=term-missing

# Check overall coverage (should remain ≥78%)
pytest tests/unit/ --cov=src --cov-report=html

5.2 Code Quality Checks

# Format (auto-fix)
black src/ tests/
isort src/ tests/

# Lint (auto-fix where possible)
ruff check src/ tests/ --fix

# Type check
mypy src/

# Security scan
bandit -r src/

# Pre-commit hooks
pre-commit run --all-files

5.3 Manual Testing with MCP Inspector

# Start MCP Inspector
uv run mcp dev src/main.py

# Open http://localhost:5173
# Test your new tool with real inputs
# Verify responses match expectations

5.4 Test on Real Hardware (if possible)

If you have access to a UniFi controller:

1. Configure local gateway API credentials
2. Test tool with real data
3. Verify API endpoint exists and returns expected format
4. Document any discrepancies in API.md

Quality Checklist

Before submitting:

Code Quality

• Tool follows naming convention ({action}_{resource})
• Comprehensive docstring with examples
• Input validation using Pydantic models
• Confirmation required for mutating operations
• Dry-run mode implemented
• Actionable error messages
• Async/await used consistently
• Type hints throughout
• No hardcoded values (use constants)
• Follows existing code patterns

Testing

• TDD: Tests written before implementation
• 85%+ coverage for new code
• Tests cover happy path
• Tests cover error scenarios
• Tests cover edge cases
• Mock API responses properly
• No integration tests without real hardware
• All tests passing

Documentation

• API.md updated with complete tool documentation
• Examples provided in docstring and API.md
• Error handling documented
• API endpoint documented
• API mode support clearly indicated
• CHANGELOG.md updated
• README.md updated (if major feature)

Integration

• Tool registered in src/main.py
• Imports organized correctly
• No circular dependencies
• Compatible with existing tools

Quality Gates

• black formatting passes
• isort import sorting passes
• ruff linting passes
• mypy type checking passes
• bandit security scan passes
• pytest all tests pass
• Coverage ≥78% overall, ≥85% for new code
• Pre-commit hooks pass
• MCP Inspector manual testing successful

Common Patterns in UniFi MCP Server

Pattern 1: List/Query Tools (Read-Only)

@mcp.tool()
async def list_resources(
    site_id: str,
    filters: dict | None = None,
    settings: Context = None,
) -> list[dict]:
    """List resources in the UniFi controller.

    This is a read-only operation requiring no confirmation.
    """
    client = settings.get("client")
    response = await client.request(
        method="GET",
        endpoint=f"/api/s/{site_id}/rest/resource",
        params=filters or {}
    )
    return response.get("data", [])

Pattern 2: Mutating Tools with Confirmation

@mcp.tool()
async def update_resource(
    site_id: str,
    resource_id: str,
    updates: dict,
    settings: Context = None,
    confirm: bool = False,
    dry_run: bool = False,
) -> dict:
    """Update a resource (requires confirmation).

    Mutating operations always require confirm=True.
    """
    if not dry_run and not confirm:
        raise PermissionError("Set confirm=True to proceed")

    if dry_run:
        return {"dry_run": True, "preview": updates}

    client = settings.get("client")
    response = await client.request(
        method="PUT",
        endpoint=f"/api/s/{site_id}/rest/resource/{resource_id}",
        json=updates
    )
    return response

Pattern 3: Tools with Pagination

@mcp.tool()
async def list_large_dataset(
    site_id: str,
    limit: int = 100,
    offset: int = 0,
    settings: Context = None,
) -> dict:
    """List resources with pagination support."""
    client = settings.get("client")
    response = await client.request(
        method="GET",
        endpoint=f"/api/s/{site_id}/rest/resource",
        params={"limit": limit, "offset": offset}
    )

    return {
        "data": response.get("data", []),
        "total": response.get("meta", {}).get("total", 0),
        "limit": limit,
        "offset": offset
    }

Pattern 4: Multi-Site Aggregation

@mcp.tool()
async def aggregate_across_sites(
    settings: Context = None,
) -> dict:
    """Aggregate data across all sites."""
    client = settings.get("client")

    # Get all sites
    sites_response = await client.request(
        method="GET",
        endpoint="/api/self/sites"
    )
    sites = sites_response.get("data", [])

    # Aggregate data
    results = []
    for site in sites:
        site_id = site["name"]
        site_data = await client.request(
            method="GET",
            endpoint=f"/api/s/{site_id}/rest/resource"
        )
        results.append({
            "site_id": site_id,
            "data": site_data.get("data", [])
        })

    return {"sites": results}

References

Project Documentation

• README.md - Project overview and quick start
• API.md - Complete MCP tool reference
• AGENTS.md - AI agent development guidelines
• DEVELOPMENT_PLAN.md - Roadmap and feature planning
• TESTING_PLAN.md - Testing strategy and coverage goals
• CONTRIBUTING.md - Contribution guidelines

UniFi API Documentation

• docs/UNIFI_API.md - Verified endpoint documentation for v10.0.156+
• UniFi Network Developer Documentation

Code Examples

• src/tools/devices.py - Device management examples
• src/tools/firewall.py - Firewall rule management patterns
• src/tools/zbf_tools.py - Zone-Based Firewall implementation
• src/tools/qos.py - QoS profile management examples
• tests/unit/test_topology.py - High-coverage test examples (95%+)

Last Updated: 2026-01-25 Maintained By: UniFi MCP Server Team