docs(mcp_server): update documentation and add developer guide for extensibility (#9533)

This commit is contained in:
Rubén De la Torre Vico
2025-12-15 12:35:59 +01:00
committed by GitHub
parent ab475bafc3
commit 4015beff20
6 changed files with 962 additions and 518 deletions

View File

@@ -0,0 +1,447 @@
---
title: 'Extending the MCP Server'
---
This guide explains how to extend the Prowler MCP Server with new tools and features.
<Info>
**New to Prowler MCP Server?** Start with the user documentation:
- [Overview](/getting-started/products/prowler-mcp) - Key capabilities, use cases, and deployment options
- [Installation](/getting-started/installation/prowler-mcp) - Install locally or use the managed server
- [Configuration](/getting-started/basic-usage/prowler-mcp) - Configure Claude Desktop, Cursor, and other MCP hosts
- [Tools Reference](/getting-started/basic-usage/prowler-mcp-tools) - Complete list of all available tools
</Info>
## Introduction
The Prowler MCP Server brings the entire Prowler ecosystem to AI assistants through the [Model Context Protocol (MCP)](https://modelcontextprotocol.io). It enables seamless integration with AI tools like Claude Desktop, Cursor, and other MCP clients.
The server follows a modular architecture with three independent sub-servers:
| Sub-Server | Auth Required | Description |
|------------|---------------|-------------|
| Prowler App | Yes | Full access to Prowler Cloud and Self-Managed features |
| Prowler Hub | No | Security checks catalog with **over 1000 checks**, fixers, and **70+ compliance frameworks** |
| Prowler Documentation | No | Full-text search and retrieval of official documentation |
<Note>
For a complete list of tools and their descriptions, see the [Tools Reference](/getting-started/basic-usage/prowler-mcp-tools).
</Note>
## Architecture Overview
The MCP Server architecture is illustrated in the [Overview documentation](/getting-started/products/prowler-mcp#mcp-server-architecture). AI assistants connect through the MCP protocol to access Prowler's three main components.
### Server Structure
The main server orchestrates three sub-servers with prefixed namespacing:
```
mcp_server/prowler_mcp_server/
├── server.py # Main orchestrator
├── main.py # CLI entry point
├── prowler_hub/
├── prowler_app/
│ ├── tools/ # Tool implementations
│ ├── models/ # Pydantic models
│ └── utils/ # API client, auth, loader
└── prowler_documentation/
```
### Tool Registration Patterns
The MCP Server uses two patterns for tool registration:
1. **Direct Decorators** (Prowler Hub/Docs): Tools are registered using `@mcp.tool()` decorators
2. **Auto-Discovery** (Prowler App): All public methods of `BaseTool` subclasses are auto-registered
## Adding Tools to Prowler App
### Step 1: Create the Tool Class
Create a new file or add to an existing file in `prowler_app/tools/`:
```python
# prowler_app/tools/new_feature.py
from typing import Any
from pydantic import Field
from prowler_mcp_server.prowler_app.models.new_feature import (
FeatureListResponse,
DetailedFeature,
)
from prowler_mcp_server.prowler_app.tools.base import BaseTool
class NewFeatureTools(BaseTool):
"""Tools for managing new features."""
async def list_features(
self,
status: str | None = Field(
default=None,
description="Filter by status (active, inactive, pending)"
),
page_size: int = Field(
default=50,
description="Number of results per page (1-100)"
),
) -> dict[str, Any]:
"""List all features with optional filtering.
Returns a lightweight list of features optimized for LLM consumption.
Use get_feature for complete information about a specific feature.
"""
# Validate parameters
self.api_client.validate_page_size(page_size)
# Build query parameters
params: dict[str, Any] = {"page[size]": page_size}
if status:
params["filter[status]"] = status
# Make API request
clean_params = self.api_client.build_filter_params(params)
response = await self.api_client.get("/api/v1/features", params=clean_params)
# Transform to LLM-friendly format
return FeatureListResponse.from_api_response(response).model_dump()
async def get_feature(
self,
feature_id: str = Field(description="The UUID of the feature"),
) -> dict[str, Any]:
"""Get detailed information about a specific feature.
Returns complete feature details including configuration and metadata.
"""
try:
response = await self.api_client.get(f"/api/v1/features/{feature_id}")
return DetailedFeature.from_api_response(response["data"]).model_dump()
except Exception as e:
self.logger.error(f"Failed to get feature {feature_id}: {e}")
return {"error": str(e), "status": "failed"}
```
### Step 2: Create the Models
Create corresponding models in `prowler_app/models/`:
```python
# prowler_app/models/new_feature.py
from typing import Any
from pydantic import Field
from prowler_mcp_server.prowler_app.models.base import MinimalSerializerMixin
class SimplifiedFeature(MinimalSerializerMixin):
"""Lightweight feature for list operations."""
id: str = Field(description="Unique feature identifier")
name: str = Field(description="Feature name")
status: str = Field(description="Current status")
@classmethod
def from_api_response(cls, data: dict[str, Any]) -> "SimplifiedFeature":
"""Transform API response to simplified format."""
attributes = data.get("attributes", {})
return cls(
id=data["id"],
name=attributes["name"],
status=attributes["status"],
)
class DetailedFeature(SimplifiedFeature):
"""Extended feature with complete details."""
description: str | None = Field(default=None, description="Feature description")
configuration: dict[str, Any] | None = Field(default=None, description="Configuration")
created_at: str = Field(description="Creation timestamp")
updated_at: str = Field(description="Last update timestamp")
@classmethod
def from_api_response(cls, data: dict[str, Any]) -> "DetailedFeature":
"""Transform API response to detailed format."""
attributes = data.get("attributes", {})
return cls(
id=data["id"],
name=attributes["name"],
status=attributes["status"],
description=attributes.get("description"),
configuration=attributes.get("configuration"),
created_at=attributes["created_at"],
updated_at=attributes["updated_at"],
)
class FeatureListResponse(MinimalSerializerMixin):
"""Response wrapper for feature list operations."""
count: int = Field(description="Total number of features")
features: list[SimplifiedFeature] = Field(description="List of features")
@classmethod
def from_api_response(cls, response: dict[str, Any]) -> "FeatureListResponse":
"""Transform API response to list format."""
data = response.get("data", [])
features = [SimplifiedFeature.from_api_response(item) for item in data]
return cls(count=len(features), features=features)
```
### Step 3: Verify Auto-Discovery
No manual registration is needed. The `tool_loader.py` automatically discovers and registers all `BaseTool` subclasses. Verify your tool is loaded by checking the server logs:
```
INFO - Auto-registered 2 tools from NewFeatureTools
INFO - Loaded and registered: NewFeatureTools
```
## Adding Tools to Prowler Hub/Docs
For Prowler Hub or Documentation tools, use the `@mcp.tool()` decorator directly:
```python
# prowler_hub/server.py
from fastmcp import FastMCP
hub_mcp_server = FastMCP("prowler-hub")
@hub_mcp_server.tool()
async def get_new_artifact(
artifact_id: str,
) -> dict:
"""Fetch a specific artifact from Prowler Hub.
Args:
artifact_id: The unique identifier of the artifact
Returns:
Dictionary containing artifact details
"""
response = prowler_hub_client.get(f"/artifact/{artifact_id}")
response.raise_for_status()
return response.json()
```
## Model Design Patterns
### MinimalSerializerMixin
All models should use `MinimalSerializerMixin` to optimize responses for LLM consumption:
```python
from prowler_mcp_server.prowler_app.models.base import MinimalSerializerMixin
class MyModel(MinimalSerializerMixin):
"""Model that excludes empty values from serialization."""
required_field: str
optional_field: str | None = None # Excluded if None
empty_list: list = [] # Excluded if empty
```
This mixin automatically excludes:
- `None` values
- Empty strings
- Empty lists
- Empty dictionaries
### Two-Tier Model Pattern
Use two-tier models for efficient responses:
- **Simplified**: Lightweight models for list operations
- **Detailed**: Extended models for single-item retrieval
```python
class SimplifiedItem(MinimalSerializerMixin):
"""Use for list operations - minimal fields."""
id: str
name: str
status: str
class DetailedItem(SimplifiedItem):
"""Use for get operations - extends simplified with details."""
description: str | None = None
configuration: dict | None = None
created_at: str
updated_at: str
```
### Factory Method Pattern
Always implement `from_api_response()` for API transformation:
```python
@classmethod
def from_api_response(cls, data: dict[str, Any]) -> "MyModel":
"""Transform API response to model.
This method handles the JSON:API format used by Prowler API,
extracting attributes and relationships as needed.
"""
attributes = data.get("attributes", {})
return cls(
id=data["id"],
name=attributes["name"],
# ... map other fields
)
```
## API Client Usage
The `ProwlerAPIClient` is a singleton that handles authentication and HTTP requests:
```python
class MyTools(BaseTool):
async def my_tool(self) -> dict:
# GET request
response = await self.api_client.get("/api/v1/endpoint", params={"key": "value"})
# POST request
response = await self.api_client.post(
"/api/v1/endpoint",
json_data={"data": {"type": "items", "attributes": {...}}}
)
# PATCH request
response = await self.api_client.patch(
f"/api/v1/endpoint/{id}",
json_data={"data": {"attributes": {...}}}
)
# DELETE request
response = await self.api_client.delete(f"/api/v1/endpoint/{id}")
```
### Helper Methods
The API client provides useful helper methods:
```python
# Validate page size (1-1000)
self.api_client.validate_page_size(page_size)
# Normalize date range with max days limit
date_range = self.api_client.normalize_date_range(date_from, date_to, max_days=2)
# Build filter parameters (handles type conversion)
clean_params = self.api_client.build_filter_params({
"filter[status]": "active",
"filter[severity__in]": ["high", "critical"], # Converts to comma-separated
"filter[muted]": True, # Converts to "true"
})
# Poll async task until completion
result = await self.api_client.poll_task_until_complete(
task_id=task_id,
timeout=60,
poll_interval=1.0
)
```
## Best Practices
### Tool Docstrings
Tool docstrings become description that is going to be read by the LLM. Provide clear usage instructions and common workflows:
```python
async def search_items(self, status: str = Field(...)) -> dict:
"""Search items with advanced filtering.
Returns a lightweight list optimized for LLM consumption.
Use get_item for complete details about a specific item.
Common workflows:
- Find critical items: status="critical"
- Find recent items: Use date_from parameter
"""
```
### Error Handling
Return structured error responses instead of raising exceptions:
```python
async def get_item(self, item_id: str) -> dict:
try:
response = await self.api_client.get(f"/api/v1/items/{item_id}")
return DetailedItem.from_api_response(response["data"]).model_dump()
except Exception as e:
self.logger.error(f"Failed to get item {item_id}: {e}")
return {"error": str(e), "status": "failed"}
```
### Parameter Descriptions
Use Pydantic `Field()` with clear descriptions. This also helps LLMs understand
the purpose of each parameter, so be as descriptive as possible:
```python
async def list_items(
self,
severity: list[str] = Field(
default=[],
description="Filter by severity levels (critical, high, medium, low)"
),
status: str | None = Field(
default=None,
description="Filter by status (PASS, FAIL, MANUAL)"
),
page_size: int = Field(
default=50,
description="Results per page"
),
) -> dict:
```
## Development Commands
```bash
# Navigate to MCP server directory
cd mcp_server
# Run in STDIO mode (default)
uv run prowler-mcp
# Run in HTTP mode
uv run prowler-mcp --transport http --host 0.0.0.0 --port 8000
# Run with environment variables
PROWLER_APP_API_KEY="pk_xxx" uv run prowler-mcp
```
For complete installation and deployment options, see:
- [Installation Guide](/getting-started/installation/prowler-mcp#from-source-development) - Development setup instructions
- [Configuration Guide](/getting-started/basic-usage/prowler-mcp) - MCP client configuration
For development I recommend to use the [Model Context Protocol Inspector](https://github.com/modelcontextprotocol/inspector) as MCP client to test and debug your tools.
## Related Documentation
<CardGroup cols={2}>
<Card title="MCP Server Overview" icon="circle-info" href="/getting-started/products/prowler-mcp">
Key capabilities, use cases, and deployment options
</Card>
<Card title="Tools Reference" icon="wrench" href="/getting-started/basic-usage/prowler-mcp-tools">
Complete reference of all available tools
</Card>
<Card title="Prowler Hub" icon="database" href="/getting-started/products/prowler-hub">
Security checks and compliance frameworks catalog
</Card>
<Card title="Lighthouse AI" icon="robot" href="/getting-started/products/prowler-lighthouse-ai">
AI-powered security analyst
</Card>
</CardGroup>
## Additional Resources
- [MCP Protocol Specification](https://modelcontextprotocol.io) - Model Context Protocol details
- [Prowler API Documentation](https://api.prowler.com/api/v1/docs) - API reference
- [Prowler Hub API](https://hub.prowler.com/api/docs) - Hub API reference
- [GitHub Repository](https://github.com/prowler-cloud/prowler) - Source code

View File

@@ -19,9 +19,7 @@
"groups": [
{
"group": "Welcome",
"pages": [
"introduction"
]
"pages": ["introduction"]
},
{
"group": "Prowler Cloud",
@@ -51,9 +49,7 @@
},
{
"group": "Prowler Lighthouse AI",
"pages": [
"getting-started/products/prowler-lighthouse-ai"
]
"pages": ["getting-started/products/prowler-lighthouse-ai"]
},
{
"group": "Prowler MCP Server",
@@ -153,9 +149,7 @@
"user-guide/cli/tutorials/quick-inventory",
{
"group": "Tutorials",
"pages": [
"user-guide/cli/tutorials/parallel-execution"
]
"pages": ["user-guide/cli/tutorials/parallel-execution"]
}
]
},
@@ -243,9 +237,7 @@
},
{
"group": "LLM",
"pages": [
"user-guide/providers/llm/getting-started-llm"
]
"pages": ["user-guide/providers/llm/getting-started-llm"]
},
{
"group": "Oracle Cloud Infrastructure",
@@ -258,9 +250,7 @@
},
{
"group": "Compliance",
"pages": [
"user-guide/compliance/tutorials/threatscore"
]
"pages": ["user-guide/compliance/tutorials/threatscore"]
}
]
},
@@ -277,7 +267,8 @@
"developer-guide/outputs",
"developer-guide/integrations",
"developer-guide/security-compliance-framework",
"developer-guide/lighthouse"
"developer-guide/lighthouse",
"developer-guide/mcp-server"
]
},
{
@@ -313,21 +304,15 @@
},
{
"tab": "Security",
"pages": [
"security"
]
"pages": ["security"]
},
{
"tab": "Contact Us",
"pages": [
"contact"
]
"pages": ["contact"]
},
{
"tab": "Troubleshooting",
"pages": [
"troubleshooting"
]
"pages": ["troubleshooting"]
},
{
"tab": "About Us",

View File

@@ -10,7 +10,7 @@ Complete reference guide for all tools available in the Prowler MCP Server. Tool
|----------|------------|------------------------|
| Prowler Hub | 10 tools | No |
| Prowler Documentation | 2 tools | No |
| Prowler Cloud/App | 28 tools | Yes |
| Prowler Cloud/App | 22 tools | Yes |
## Tool Naming Convention
@@ -20,6 +20,66 @@ All tools follow a consistent naming pattern with prefixes:
- `prowler_docs_*` - Prowler documentation search and retrieval
- `prowler_app_*` - Prowler Cloud and App (Self-Managed) management tools
## Prowler Cloud/App Tools
Manage Prowler Cloud or Prowler App (Self-Managed) features. **Requires authentication.**
<Note>
These tools require a valid API key. See the [Configuration Guide](/getting-started/basic-usage/prowler-mcp) for authentication setup.
</Note>
### Findings Management
Tools for searching, viewing, and analyzing security findings across all cloud providers.
- **`prowler_app_search_security_findings`** - Search and filter security findings with advanced filtering options (severity, status, provider, region, service, check ID, date range, muted status)
- **`prowler_app_get_finding_details`** - Get comprehensive details about a specific finding including remediation guidance, check metadata, and resource relationships
- **`prowler_app_get_findings_overview`** - Get aggregate statistics and trends about security findings as a markdown report
### Provider Management
Tools for managing cloud provider connections in Prowler.
- **`prowler_app_search_providers`** - Search and view configured providers with their connection status
- **`prowler_app_connect_provider`** - Register and connect a provider with credentials for security scanning
- **`prowler_app_delete_provider`** - Permanently remove a provider from Prowler
### Scan Management
Tools for managing and monitoring security scans.
- **`prowler_app_list_scans`** - List and filter security scans across all providers
- **`prowler_app_get_scan`** - Get comprehensive details about a specific scan (progress, duration, resource counts)
- **`prowler_app_trigger_scan`** - Trigger a manual security scan for a provider
- **`prowler_app_schedule_daily_scan`** - Schedule automated daily scans for continuous monitoring
- **`prowler_app_update_scan`** - Update scan name for better organization
### Resources Management
Tools for searching, viewing, and analyzing cloud resources discovered by Prowler.
- **`prowler_app_list_resources`** - List and filter cloud resources with advanced filtering options (provider, region, service, resource type, tags)
- **`prowler_app_get_resource`** - Get comprehensive details about a specific resource including configuration, metadata, and finding relationships
- **`prowler_app_get_resources_overview`** - Get aggregate statistics about cloud resources as a markdown report
### Muting Management
Tools for managing finding muting, including pattern-based bulk muting (mutelist) and finding-specific mute rules.
#### Mutelist (Pattern-Based Muting)
- **`prowler_app_get_mutelist`** - Retrieve the current mutelist configuration for the tenant
- **`prowler_app_set_mutelist`** - Create or update the mutelist configuration for pattern-based bulk muting
- **`prowler_app_delete_mutelist`** - Remove the mutelist configuration from the tenant
#### Mute Rules (Finding-Specific Muting)
- **`prowler_app_list_mute_rules`** - Search and filter mute rules with pagination support
- **`prowler_app_get_mute_rule`** - Retrieve comprehensive details about a specific mute rule
- **`prowler_app_create_mute_rule`** - Create a new mute rule to mute specific findings with documentation and audit trail
- **`prowler_app_update_mute_rule`** - Update a mute rule's name, reason, or enabled status
- **`prowler_app_delete_mute_rule`** - Delete a mute rule from the system
## Prowler Hub Tools
Access Prowler's security check catalog and compliance frameworks. **No authentication required.**
@@ -53,60 +113,6 @@ Search and access official Prowler documentation. **No authentication required.*
- **`prowler_docs_search`** - Search the official Prowler documentation using full-text search
- **`prowler_docs_get_document`** - Retrieve the full markdown content of a specific documentation file
## Prowler Cloud/App Tools
Manage Prowler Cloud or Prowler App (Self-Managed) features. **Requires authentication.**
<Note>
These tools require a valid API key. See the [Configuration Guide](/getting-started/basic-usage/prowler-mcp) for authentication setup.
</Note>
### Findings Management
- **`prowler_app_list_findings`** - List security findings with advanced filtering
- **`prowler_app_get_finding`** - Get detailed information about a specific finding
- **`prowler_app_get_latest_findings`** - Retrieve latest findings from the most recent scans
- **`prowler_app_get_findings_metadata`** - Get unique metadata values from filtered findings
- **`prowler_app_get_latest_findings_metadata`** - Get metadata from latest findings across all providers
### Provider Management
- **`prowler_app_list_providers`** - List all providers with filtering options
- **`prowler_app_create_provider`** - Create a new provider in the current tenant
- **`prowler_app_get_provider`** - Get detailed information about a specific provider
- **`prowler_app_update_provider`** - Update provider details (alias, etc.)
- **`prowler_app_delete_provider`** - Delete a specific provider
- **`prowler_app_test_provider_connection`** - Test provider connection status
### Provider Secrets Management
- **`prowler_app_list_provider_secrets`** - List all provider secrets with filtering
- **`prowler_app_add_provider_secret`** - Add or update credentials for a provider
- **`prowler_app_get_provider_secret`** - Get detailed information about a provider secret
- **`prowler_app_update_provider_secret`** - Update provider secret details
- **`prowler_app_delete_provider_secret`** - Delete a provider secret
### Scan Management
- **`prowler_app_list_scans`** - List all scans with filtering options
- **`prowler_app_create_scan`** - Trigger a manual scan for a specific provider
- **`prowler_app_get_scan`** - Get detailed information about a specific scan
- **`prowler_app_update_scan`** - Update scan details
- **`prowler_app_get_scan_compliance_report`** - Download compliance report as CSV
- **`prowler_app_get_scan_report`** - Download ZIP file containing complete scan report
### Schedule Management
- **`prowler_app_schedules_daily_scan`** - Create a daily scheduled scan for a provider
### Processor Management
- **`prowler_app_processors_list`** - List all processors with filtering
- **`prowler_app_processors_create`** - Create a new processor (currently only mute lists supported)
- **`prowler_app_processors_retrieve`** - Get processor details by ID
- **`prowler_app_processors_partial_update`** - Update processor configuration
- **`prowler_app_processors_destroy`** - Delete a processor
## Usage Tips
- Use natural language to interact with the tools through your AI assistant

View File

@@ -19,12 +19,11 @@ The Prowler MCP Server provides three main integration points:
### 1. Prowler Cloud and Prowler App (Self-Managed)
Full access to Prowler Cloud platform and self-managed Prowler App for:
- **Provider Management**: Create, configure, and manage cloud providers (AWS, Azure, GCP, etc.).
- **Scan Orchestration**: Trigger on-demand scans and schedule recurring security assessments.
- **Findings Analysis**: Query, filter, and analyze security findings across all your cloud environments.
- **Compliance Reporting**: Generate compliance reports for various frameworks (CIS, PCI-DSS, HIPAA, etc.).
- **Secrets Management**: Securely manage provider credentials and connection details.
- **Processor Configuration**: Set up the [Prowler Mutelist](/user-guide/tutorials/prowler-app-mute-findings) to mute findings.
- **Findings Analysis**: Query, filter, and analyze security findings across all your cloud environments
- **Provider Management**: Create, configure, and manage your configured Prowler providers (AWS, Azure, GCP, etc.)
- **Scan Orchestration**: Trigger on-demand scans and schedule recurring security assessments
- **Resource Inventory**: Search and view detailed information about your audited resources
- **Muting Management**: Create and manage muting lists/rules to suppress non-relevant findings
### 2. Prowler Hub
@@ -49,7 +48,10 @@ The following diagram illustrates the Prowler MCP Server architecture and its in
<img className="block dark:hidden" src="/images/prowler_mcp_schema_light.png" alt="Prowler MCP Server Schema" />
<img className="hidden dark:block" src="/images/prowler_mcp_schema_dark.png" alt="Prowler MCP Server Schema" />
The architecture shows how AI assistants connect through the MCP protocol to access Prowler's three main components: Prowler Cloud/App for security operations, Prowler Hub for security knowledge, and Prowler Documentation for guidance and reference.
The architecture shows how AI assistants connect through the MCP protocol to access Prowler's three main components:
- Prowler Cloud/App for security operations
- Prowler Hub for security knowledge
- Prowler Documentation for guidance and reference.
## Use Cases
@@ -57,12 +59,12 @@ The Prowler MCP Server enables powerful workflows through AI assistants:
**Security Operations**
- "Show me all critical findings from my AWS production accounts"
- "What is my compliance status for the PCI standards accross all my AWS accounts according to the latest Prowler scan results?"
- "Register my new AWS account in Prowler and run an scheduled scan every day"
- "Register my new AWS account in Prowler and run a scheduled scan every day"
- "List all muted findings and detect what findgings are muted by a not enough good reason in relation to their severity"
**Security Research**
- "Explain what the S3 bucket public access check does"
- "Find all checks related to encryption at rest"
- "Explain what the S3 bucket public access Prowler check does"
- "Find all Prowler checks related to encryption at rest"
- "What is the latest version of the CIS that Prowler is covering per provider?"
**Documentation & Learning**

310
mcp_server/AGENTS.md Normal file
View File

@@ -0,0 +1,310 @@
# Prowler MCP Server - AI Agent Ruleset
**Complete guide for AI agents and developers working on the Prowler MCP Server - the Model Context Protocol server that provides AI agents access to the Prowler ecosystem.**
## Project Overview
The Prowler MCP Server brings the entire Prowler ecosystem to AI assistants through
the Model Context Protocol (MCP). It enables seamless integration with AI tools
like Claude Desktop, Cursor, and other MCP hosts, allowing interaction with
Prowler's security capabilities through natural language.
---
## Critical Rules
### Tool Implementation
- **ALWAYS**: Extend `BaseTool` ABC for new Prowler App tools (auto-registration)
- **ALWAYS**: Use `@mcp.tool()` decorator for Hub/Docs tools (manual registration)
- **NEVER**: Manually register BaseTool subclasses (auto-discovered via `load_all_tools()`)
- **NEVER**: Import tools directly in server.py (tool_loader handles discovery)
### Models
- **ALWAYS**: Use `MinimalSerializerMixin` for LLM-optimized responses
- **ALWAYS**: Implement `from_api_response()` factory method for API transformations
- **ALWAYS**: Use two-tier models (Simplified for lists, Detailed for single items)
- **NEVER**: Return raw API responses (transform to simplified models)
### API Client
- **ALWAYS**: Use singleton `ProwlerAPIClient` via `self.api_client` in tools
- **ALWAYS**: Use `build_filter_params()` for query parameter normalization
- **NEVER**: Create new httpx clients in tools (use shared client)
---
## Architecture
### Three Sub-Servers Pattern
The main server (`server.py`) orchestrates three independent sub-servers with prefixed tool namespacing:
```python
# server.py imports sub-servers with prefixes
await prowler_mcp_server.import_server(hub_mcp_server, prefix="prowler_hub")
await prowler_mcp_server.import_server(app_mcp_server, prefix="prowler_app")
await prowler_mcp_server.import_server(docs_mcp_server, prefix="prowler_docs")
```
This pattern ensures:
- Failures in one sub-server do not block others
- Clear tool namespacing for LLM disambiguation
- Independent development and testing
### Tool Naming Convention
All tools follow a consistent naming pattern with prefixes:
- `prowler_hub_*` - Prowler Hub catalog and compliance tools
- `prowler_docs_*` - Prowler documentation search and retrieval
- `prowler_app_*` - Prowler Cloud and App (Self-Managed) management tools
### Tool Registration Patterns
**Pattern 1: Prowler Hub/Docs (Direct Decorators)**
```python
# prowler_hub/server.py or prowler_documentation/server.py
hub_mcp_server = FastMCP("prowler-hub")
@hub_mcp_server.tool()
async def get_checks(providers: str | None = None) -> dict:
"""Tool docstring becomes LLM description."""
# Direct implementation
response = prowler_hub_client.get("/check", params=params)
return response.json()
```
**Pattern 2: Prowler App (BaseTool Auto-Registration)**
```python
# prowler_app/tools/findings.py
class FindingsTools(BaseTool):
async def search_security_findings(
self,
severity: list[str] = Field(default=[], description="Filter by severity")
) -> dict:
"""Docstring becomes LLM description."""
response = await self.api_client.get("/api/v1/findings")
return SimplifiedFinding.from_api_response(response).model_dump()
```
NOTE: Only public methods of `BaseTool` subclasses are registered as tools.
---
## Tech Stack
- **Language**: Python 3.12+
- **MCP Framework**: FastMCP 2.13.1
- **HTTP Client**: httpx (async)
- **Validation**: Pydantic with MinimalSerializerMixin
- **Package Manager**: uv
---
## Project Structure
```
mcp_server/
├── README.md # User documentation
├── AGENTS.md # This file - AI agent guidelines
├── CHANGELOG.md # Version history
├── pyproject.toml # Project metadata and dependencies
├── Dockerfile # Container image definition
├── entrypoint.sh # Docker entrypoint script
└── prowler_mcp_server/
├── __init__.py # Version info
├── main.py # CLI entry point
├── server.py # Main FastMCP server orchestration
├── lib/
│ └── logger.py # Structured logging
├── prowler_hub/
│ └── server.py # Hub tools (10 tools, no auth)
├── prowler_app/
│ ├── server.py # App server initialization
│ ├── tools/
│ │ ├── base.py # BaseTool abstract class
│ │ ├── findings.py # Findings tools
│ │ ├── providers.py # Provider tools
│ │ ├── scans.py # Scan tools
│ │ ├── resources.py # Resource tools
│ │ └── muting.py # Muting tools
│ ├── models/
│ │ ├── base.py # MinimalSerializerMixin
│ │ ├── findings.py # Finding models
│ │ ├── providers.py # Provider models
│ │ ├── scans.py # Scan models
│ │ ├── resources.py # Resource models
│ │ └── muting.py # Muting models
│ └── utils/
│ ├── api_client.py # ProwlerAPIClient singleton
│ ├── auth.py # ProwlerAppAuth (STDIO/HTTP)
│ └── tool_loader.py # Auto-discovery and registration
└── prowler_documentation/
├── server.py # Documentation tools (2 tools, no auth)
└── search_engine.py # Mintlify API integration
```
---
## Commands
NOTE: To run a python command always use `uv run <command>` from within the `mcp_server/` directory.
### Development
```bash
# Navigate to MCP server directory
cd mcp_server
# Run in STDIO mode (default)
uv run prowler-mcp
# Run in HTTP mode
uv run prowler-mcp --transport http --host 0.0.0.0 --port 8000
# Run from anywhere using uvx
uvx /path/to/prowler/mcp_server/
```
---
## Development Patterns
### Adding New Tools to Prowler App
1. **Create or extend a tool class** in `prowler_app/tools/`:
```python
# prowler_app/tools/new_feature.py
from pydantic import Field
from prowler_mcp_server.prowler_app.tools.base import BaseTool
from prowler_mcp_server.prowler_app.models.new_feature import FeatureResponse
class NewFeatureTools(BaseTool):
async def list_features(
self,
status: str | None = Field(default=None, description="Filter by status")
) -> dict:
"""List all features with optional filtering.
Returns a simplified list of features optimized for LLM consumption.
"""
params = {}
if status:
params["filter[status]"] = status
clean_params = self.api_client.build_filter_params(params)
response = await self.api_client.get("/api/v1/features", params=clean_params)
return FeatureResponse.from_api_response(response).model_dump()
```
2. **Create corresponding models** in `prowler_app/models/`:
```python
# prowler_app/models/new_feature.py
from pydantic import Field
from prowler_mcp_server.prowler_app.models.base import MinimalSerializerMixin
class SimplifiedFeature(MinimalSerializerMixin):
"""Lightweight feature for list operations."""
id: str
name: str
status: str
class DetailedFeature(SimplifiedFeature):
"""Extended feature with complete details."""
description: str | None = None
created_at: str
updated_at: str
@classmethod
def from_api_response(cls, data: dict) -> "DetailedFeature":
"""Transform API response to model."""
attributes = data.get("attributes", {})
return cls(
id=data["id"],
name=attributes["name"],
status=attributes["status"],
description=attributes.get("description"),
created_at=attributes["created_at"],
updated_at=attributes["updated_at"],
)
```
3. **No registration needed** - the tool loader auto-discovers BaseTool subclasses
### Adding Tools to Prowler Hub/Docs
Use the `@mcp.tool()` decorator directly:
```python
# prowler_hub/server.py
@hub_mcp_server.tool()
async def new_hub_tool(param: str) -> dict:
"""Tool description for LLM."""
response = prowler_hub_client.get("/endpoint")
return response.json()
```
---
## Code Quality Standards
### Tool Docstrings
Tool docstrings become AI agent descriptions. Write them in a clear, concise manner focusing on LLM-relevant behavior:
```python
async def search_security_findings(
self,
severity: list[str] = Field(default=[], description="Filter by severity levels")
) -> dict:
"""Search security findings with advanced filtering.
Returns a lightweight list of findings optimized for LLM consumption.
Use get_finding_details for complete information about a specific finding.
"""
```
### Model Design
- Use `MinimalSerializerMixin` to exclude None/empty values
- Implement `from_api_response()` for consistent API transformation
- Create two-tier models: Simplified (lists) and Detailed (single items)
### Error Handling
Return structured error responses rather than raising exceptions:
```python
try:
response = await self.api_client.get(f"/api/v1/items/{item_id}")
return DetailedItem.from_api_response(response["data"]).model_dump()
except Exception as e:
self.logger.error(f"Failed to get item {item_id}: {e}")
return {"error": str(e), "status": "failed"}
```
---
## QA Checklist Before Commit
- [ ] Tool docstrings are clear and describe LLM-relevant behavior
- [ ] Models use `MinimalSerializerMixin` for LLM optimization
- [ ] API responses are transformed to simplified models
- [ ] No hardcoded secrets or API keys
- [ ] Error handling returns structured responses
- [ ] New tools are auto-discovered (BaseTool subclass) or properly decorated
- [ ] Parameter descriptions use Pydantic `Field()` with clear descriptions
---
## References
- **Root Project Guide**: `../AGENTS.md`
- **FastMCP Documentation**: https://gofastmcp.com/llms.txt
- **Prowler API Documentation**: https://api.prowler.com/api/v1/docs

View File

@@ -1,18 +1,59 @@
# Prowler MCP Server
> ⚠️ **Preview Feature**: This MCP server is currently in preview and under active development. Features and functionality may change. We welcome your feedback—please report any issues on [GitHub](https://github.com/prowler-cloud/prowler/issues) or join our [Slack community](https://goto.prowler.com/slack) to discuss and share your thoughts.
**Prowler MCP Server** brings the entire Prowler ecosystem to AI assistants through the [Model Context Protocol (MCP)](https://modelcontextprotocol.io). It enables seamless integration with AI tools like Claude Desktop, Cursor, and other MCP clients, allowing interaction with Prowler's security capabilities through natural language.
Access the entire Prowler ecosystem through the Model Context Protocol (MCP). This server provides three main capabilities:
> **Preview Feature**: This MCP server is currently under active development. Features and functionality may change. We welcome your feedback—please report any issues on [GitHub](https://github.com/prowler-cloud/prowler/issues) or join our [Slack community](https://goto.prowler.com/slack).
- **Prowler Cloud and Prowler App (Self-Managed)**: Full access to Prowler Cloud platform and Prowler Self-Managed for managing providers, running scans, and analyzing security findings
- **Prowler Hub**: Access to Prowler's security checks, fixers, and compliance frameworks catalog
- **Prowler Documentation**: Search and retrieve official Prowler documentation
## Key Capabilities
## Quick Start with Hosted Server (Recommended)
### Prowler Cloud and Prowler App (Self-Managed)
**The easiest way to use Prowler MCP is through our hosted server at `https://mcp.prowler.com/mcp`**
Full access to Prowler Cloud platform and self-managed Prowler App for:
- **Findings Analysis**: Query, filter, and analyze security findings across all your cloud environments
- **Provider Management**: Create, configure, and manage your configured Prowler providers (AWS, Azure, GCP, etc.)
- **Scan Orchestration**: Trigger on-demand scans and schedule recurring security assessments
- **Resource Inventory**: Search and view detailed information about your audited resources
- **Muting Management**: Create and manage muting rules to suppress non-critical findings
No installation required! Just configure your MCP client:
### Prowler Hub
Access to Prowler's comprehensive security knowledge base:
- **Security Checks Catalog**: Browse and search **over 1000 security checks** across multiple Prowler providers
- **Check Implementation**: View the Python code that powers each security check
- **Automated Fixers**: Access remediation scripts for common security issues
- **Compliance Frameworks**: Explore mappings to **over 70 compliance standards and frameworks**
- **Provider Services**: View available services and checks for each cloud provider
### Prowler Documentation
Search and retrieve official Prowler documentation:
- **Intelligent Search**: Full-text search across all Prowler documentation
- **Contextual Results**: Get relevant documentation pages with highlighted snippets
- **Document Retrieval**: Access complete markdown content of any documentation file
## Documentation
For comprehensive guides and tutorials, see the official documentation:
| Guide | Description |
|-------|-------------|
| [Overview](https://docs.prowler.com/getting-started/products/prowler-mcp) | Key capabilities, use cases, and deployment options |
| [Installation](https://docs.prowler.com/getting-started/installation/prowler-mcp) | Docker, PyPI, and source installation |
| [Configuration](https://docs.prowler.com/getting-started/basic-usage/prowler-mcp) | Configure Claude Desktop, Cursor, and other MCP clients |
| [Tools Reference](https://docs.prowler.com/getting-started/basic-usage/prowler-mcp-tools) | Complete reference of all tools |
| [Developer Guide](https://docs.prowler.com/developer-guide/mcp-server) | How to extend with new tools |
## Deployment Options
Prowler MCP Server can be used in three ways:
### 1. Prowler Cloud MCP Server (Recommended)
**Use Prowler's managed MCP server at `https://mcp.prowler.com/mcp`**
- No installation required
- Managed and maintained by Prowler team
- Always up-to-date
```json
{
@@ -30,70 +71,37 @@ No installation required! Just configure your MCP client:
}
```
**Configuration file locations:**
- **Claude Desktop (macOS)**: `~/Library/Application Support/Claude/claude_desktop_config.json`
- **Claude Desktop (Windows)**: `%AppData%\Claude\claude_desktop_config.json`
- **Cursor**: `~/.cursor/mcp.json`
### 2. Local STDIO Mode
Get your API key at [Prowler Cloud](https://cloud.prowler.com) → Settings → API Keys
**Run the server locally on your machine**
> **Benefits:** Always up-to-date, no maintenance, managed by Prowler team
- Runs as a subprocess of your MCP client
- Requires Python 3.12+ or Docker
## Local/Self-Hosted Installation
### 3. Self-Hosted HTTP Mode
If you need to run the MCP server locally or self-host it, choose one of the following installation methods. **Configuration is the same** for both managed and local installations - just point to your local server URL instead of `https://mcp.prowler.com/mcp`.
**Deploy your own remote MCP server**
### Requirements
- Full control over deployment
- Requires Python 3.12+ or Docker
- Python 3.12+ (for source/PyPI installation)
- Docker (for Docker installation)
- Network access to `https://hub.prowler.com` (for Prowler Hub)
- Network access to `https://prowler.mintlify.app` (for Prowler Documentation)
- Network access to Prowler Cloud and Prowler App (Self-Managed) API (optional, only for Prowler Cloud/App features)
- Prowler Cloud account credentials (only for Prowler Cloud and Prowler App features)
See the [Installation Guide](https://docs.prowler.com/getting-started/installation/prowler-mcp) for complete instructions.
### Installation Methods
## Quick Installation
#### Option 1: Docker Hub (Recommended)
Pull the official image from Docker Hub:
### Docker (Recommended)
```bash
docker pull prowlercloud/prowler-mcp
```
Run in STDIO mode:
```bash
# STDIO mode
docker run --rm -i prowlercloud/prowler-mcp
# HTTP mode
docker run --rm -p 8000:8000 prowlercloud/prowler-mcp --transport http --host 0.0.0.0 --port 8000
```
Run in HTTP mode:
```bash
docker run --rm -p 8000:8000 \
prowlercloud/prowler-mcp \
--transport http --host 0.0.0.0 --port 8000
```
With environment variables:
```bash
docker run --rm -i \
-e PROWLER_APP_API_KEY="pk_your_api_key" \
-e API_BASE_URL="https://api.prowler.com/api/v1" \
prowlercloud/prowler-mcp
```
**Docker Hub:** [prowlercloud/prowler-mcp](https://hub.docker.com/r/prowlercloud/prowler-mcp)
#### Option 2: PyPI Package (Coming Soon)
```bash
pip install prowler-mcp-server
prowler-mcp --help
```
#### Option 3: From Source (Development)
Clone the repository and use `uv`:
### From Source
```bash
git clone https://github.com/prowler-cloud/prowler.git
@@ -101,400 +109,86 @@ cd prowler/mcp_server
uv run prowler-mcp --help
```
Install [uv](https://docs.astral.sh/uv/) first if needed.
#### Option 4: Build Docker Image from Source
```bash
git clone https://github.com/prowler-cloud/prowler.git
cd prowler/mcp_server
docker build -t prowler-mcp .
docker run --rm -i prowler-mcp
```
## Running
The Prowler MCP server supports two transport modes:
- **STDIO mode** (default): For direct integration with MCP clients like Claude Desktop
- **HTTP mode**: For remote access over HTTP with Bearer token authentication
### Transport Modes
#### STDIO Mode (Default)
STDIO mode is the standard MCP transport for direct client integration:
```bash
cd prowler/mcp_server
uv run prowler-mcp
# or
uv run prowler-mcp --transport stdio
```
#### HTTP Mode (Remote Server)
HTTP mode allows the server to run as a remote service accessible over HTTP:
```bash
cd prowler/mcp_server
# Run on default host and port (127.0.0.1:8000)
uv run prowler-mcp --transport http
# Run on custom host and port
uv run prowler-mcp --transport http --host 0.0.0.0 --port 8080
```
For self-deployed MCP remote server, you can use also configure the server to use a custom API base URL with the environment variable `API_BASE_URL`; and the transport mode with the environment variable `PROWLER_MCP_TRANSPORT_MODE`.
```bash
export API_BASE_URL="https://api.prowler.com/api/v1"
export PROWLER_MCP_TRANSPORT_MODE="http"
```
### Using uv directly
After installation, start the MCP server via the console script:
```bash
cd prowler/mcp_server
uv run prowler-mcp
```
Alternatively, you can run from wherever you want using `uvx` command:
```bash
uvx /path/to/prowler/mcp_server/
```
### Using Docker
#### STDIO Mode (Default)
Run the pre-built Docker container in STDIO mode:
```bash
cd prowler/mcp_server
docker run --rm --env-file ./.env -it prowler-mcp
```
#### HTTP Mode (Remote Server)
Run as a remote HTTP server:
```bash
cd prowler/mcp_server
# Run on port 8000 (accessible from host)
docker run --rm --env-file ./.env -p 8000:8000 -it prowler-mcp --transport http --host 0.0.0.0 --port 8000
# Run on custom port
docker run --rm --env-file ./.env -p 8080:8080 -it prowler-mcp --transport http --host 0.0.0.0 --port 8080
```
## Production Deployment
For production deployments that require customization, it is recommended to use the ASGI application that can be found in `prowler_mcp_server.server`. This can be run with uvicorn:
```bash
uvicorn prowler_mcp_server.server:app --host 0.0.0.0 --port 8000
```
For more details on production deployment options, see the [FastMCP production deployment guide](https://gofastmcp.com/deployment/http#production-deployment) and [uvicorn settings](https://www.uvicorn.org/settings/).
## Command Line Arguments
The Prowler MCP server supports the following command line arguments:
```
prowler-mcp [--transport {stdio,http}] [--host HOST] [--port PORT]
```
**Arguments:**
- `--transport {stdio,http}`: Transport method (default: stdio)
- `stdio`: Standard input/output transport for direct MCP client integration
- `http`: HTTP transport for remote server access
- `--host HOST`: Host to bind to for HTTP transport (default: 127.0.0.1)
- `--port PORT`: Port to bind to for HTTP transport (default: 8000)
**Examples:**
```bash
# Default STDIO mode
prowler-mcp
# Explicit STDIO mode
prowler-mcp --transport stdio
# HTTP mode with default host and port (127.0.0.1:8000)
prowler-mcp --transport http
# HTTP mode accessible from any network interface
prowler-mcp --transport http --host 0.0.0.0
# HTTP mode with custom port
prowler-mcp --transport http --host 0.0.0.0 --port 8080
```
## Available Tools
### Prowler Hub
For complete tool descriptions and parameters, see the [Tools Reference](https://docs.prowler.com/getting-started/basic-usage/prowler-mcp-tools).
All tools are exposed under the `prowler_hub` prefix.
### Tool Naming Convention
- `prowler_hub_get_check_filters`: Return available filter values for checks (providers, services, severities, categories, compliances). Call this before `prowler_hub_get_checks` to build valid queries.
- `prowler_hub_get_checks`: List checks with option of advanced filtering.
- `prowler_hub_get_check_raw_metadata`: Fetch raw check metadata JSON (low-level version of get_checks).
- `prowler_hub_get_check_code`: Fetch check implementation Python code from Prowler.
- `prowler_hub_get_check_fixer`: Fetch check fixer Python code from Prowler (if it exists).
- `prowler_hub_search_checks`: Fulltext search across check metadata.
- `prowler_hub_get_compliance_frameworks`: List/filter compliance frameworks.
- `prowler_hub_search_compliance_frameworks`: Full-text search across frameworks.
- `prowler_hub_list_providers`: List Prowler official providers and their services.
- `prowler_hub_get_artifacts_count`: Return total artifact count (checks + frameworks).
All tools follow a consistent naming pattern with prefixes:
- `prowler_app_*` - Prowler Cloud and App (Self-Managed) management tools
- `prowler_hub_*` - Prowler Hub catalog and compliance tools
- `prowler_docs_*` - Prowler documentation search and retrieval
### Prowler Documentation
## Architecture
All tools are exposed under the `prowler_docs` prefix.
- `prowler_docs_search`: Search the official Prowler documentation using fulltext search. Returns relevant documentation pages with highlighted snippets and relevance scores.
- `prowler_docs_get_document`: Retrieve the full markdown content of a specific documentation file using the path from search results.
### Prowler Cloud and Prowler App (Self-Managed)
All tools are exposed under the `prowler_app` prefix.
#### Findings Management
- `prowler_app_list_findings`: List security findings from Prowler scans with advanced filtering
- `prowler_app_get_finding`: Get detailed information about a specific security finding
- `prowler_app_get_latest_findings`: Retrieve latest findings from the latest scans for each provider
- `prowler_app_get_findings_metadata`: Fetch unique metadata values from filtered findings
- `prowler_app_get_latest_findings_metadata`: Fetch metadata from latest findings across all providers
#### Provider Management
- `prowler_app_list_providers`: List all providers with filtering options
- `prowler_app_create_provider`: Create a new provider in the current tenant
- `prowler_app_get_provider`: Get detailed information about a specific provider
- `prowler_app_update_provider`: Update provider details (alias, etc.)
- `prowler_app_delete_provider`: Delete a specific provider
- `prowler_app_test_provider_connection`: Test provider connection status
#### Provider Secrets Management
- `prowler_app_list_provider_secrets`: List all provider secrets with filtering
- `prowler_app_add_provider_secret`: Add or update credentials for a provider
- `prowler_app_get_provider_secret`: Get detailed information about a provider secret
- `prowler_app_update_provider_secret`: Update provider secret details
- `prowler_app_delete_provider_secret`: Delete a provider secret
#### Scan Management
- `prowler_app_list_scans`: List all scans with filtering options
- `prowler_app_create_scan`: Trigger a manual scan for a specific provider
- `prowler_app_get_scan`: Get detailed information about a specific scan
- `prowler_app_update_scan`: Update scan details
- `prowler_app_get_scan_compliance_report`: Download compliance report as CSV
- `prowler_app_get_scan_report`: Download ZIP file containing scan report
#### Schedule Management
- `prowler_app_schedules_daily_scan`: Create a daily scheduled scan for a provider
#### Processor Management
- `prowler_app_processors_list`: List all processors with filtering
- `prowler_app_processors_create`: Create a new processor. For now, only mute lists are supported.
- `prowler_app_processors_retrieve`: Get processor details by ID
- `prowler_app_processors_partial_update`: Update processor configuration
- `prowler_app_processors_destroy`: Delete a processor
## Configuration
### Prowler Cloud and Prowler App (Self-Managed) Authentication
> [!IMPORTANT]
> Authentication is not needed for using Prowler Hub or Prowler Documentation features.
The Prowler MCP server supports different authentication in Prowler Cloud and Prowler App (Self-Managed) methods depending on the transport mode:
#### STDIO Mode Authentication
For STDIO mode, authentication is handled via environment variables using an API key:
```bash
# Required for Prowler Cloud and Prowler App (Self-Managed) authentication
export PROWLER_APP_API_KEY="pk_your_api_key_here"
# Optional - for custom API endpoint, in case not provided Prowler Cloud API will be used
export API_BASE_URL="https://api.prowler.com/api/v1"
```
prowler_mcp_server/
├── server.py # Main orchestrator (imports sub-servers with prefixes)
├── main.py # CLI entry point
├── prowler_hub/ # tools - no authentication required
├── prowler_app/ # tools - authentication required
│ ├── tools/ # Tool implementations
│ ├── models/ # Pydantic models for LLM-optimized responses
│ └── utils/ # API client, authentication, tool loader
└── prowler_documentation/ # tools - no authentication required
```
#### HTTP Mode Authentication
**Key Features:**
- **Modular Design**: Three independent sub-servers with prefixed namespacing
- **Auto-Discovery**: Prowler App tools are automatically discovered and registered
- **LLM Optimization**: Response models minimize token usage by excluding empty values
- **Dual Transport**: Supports both STDIO (local) and HTTP (remote) modes
For HTTP mode (remote server), authentication is handled via Bearer tokens. The MCP server supports both JWT tokens and API keys:
## Use Cases
**Option 1: Using API Keys (Recommended)**
Use your Prowler API key directly in the MCP client configuration with Bearer token format:
```
Authorization: Bearer pk_your_api_key_here
```
The Prowler MCP Server enables powerful workflows through AI assistants:
**Option 2: Using JWT Tokens**
You need to obtain a JWT token from Prowler Cloud/App and include the generated token in the MCP client configuration. To get a valid token, you can use the following command (replace the email and password with your own credentials):
**Security Operations**
- "Show me all critical findings from my AWS production accounts"
- "Register my new AWS account in Prowler and run a scheduled scan every day"
- "List all muted findings and detect what findgings are muted by a not enough good reason in relation to their severity"
```bash
curl -X POST https://api.prowler.com/api/v1/tokens \
-H "Content-Type: application/vnd.api+json" \
-H "Accept: application/vnd.api+json" \
-d '{
"data": {
"type": "tokens",
"attributes": {
"email": "your-email@example.com",
"password": "your-password"
}
}
}'
```
**Security Research**
- "Explain what the S3 bucket public access Prowler check does"
- "Find all Prowler checks related to encryption at rest"
- "What is the latest version of the CIS that Prowler is covering per provider?"
The response will be a JWT token that you can use to [authenticate your MCP client](#http-mode-configuration-remote-server).
**Documentation & Learning**
- "How do I configure Prowler to scan my GCP organization?"
- "What authentication methods does Prowler support for Azure?"
- "How can I contribute with a new security check to Prowler?"
### MCP Client Configuration
## Requirements
Configure your MCP client, like Claude Desktop, Cursor, etc, to connect to the server. The configuration depends on whether you're running in STDIO mode (local) or HTTP mode (remote).
**For Prowler Cloud MCP Server:**
- Prowler Cloud account and API key (only for Prowler Cloud/App features)
#### STDIO Mode Configuration
**For self-hosted STDIO/HTTP Mode:**
- Python 3.12+ or Docker
- Network access to:
- `https://hub.prowler.com` (for Prowler Hub)
- `https://docs.prowler.com` (for Prowler Documentation)
- Prowler Cloud API or self-hosted Prowler App API (for Prowler Cloud/App features)
For local execution, configure your MCP client to launch the server directly. Below are examples for both direct execution and Docker deployment; consult your client's documentation for exact locations.
> **No Authentication Required**: Prowler Hub and Prowler Documentation features work without authentication. A Prowler API key is only required to access Prowler Cloud or Prowler App (Self-Managed) features.
##### Using uvx (Direct Execution)
## Configuring MCP Hosts
```json
{
"mcpServers": {
"prowler": {
"command": "uvx",
"args": ["/path/to/prowler/mcp_server/"],
"env": {
"PROWLER_APP_API_KEY": "pk_your_api_key_here",
"API_BASE_URL": "https://api.prowler.com/api/v1" // Optional, in case not provided Prowler Cloud API will be used
}
}
}
}
```
To configure your MCP host (Claude Code, Cursor, etc.) see the [Configuration Guide](https://docs.prowler.com/getting-started/basic-usage/prowler-mcp) for detailed setup instructions.
##### Using Docker
## Contributing
```json
{
"mcpServers": {
"prowler": {
"command": "docker",
"args": [
"run", "--rm", "-i",
"--env", "PROWLER_APP_API_KEY=pk_your_api_key_here",
"--env", "API_BASE_URL=https://api.prowler.com/api/v1", // Optional, in case not provided Prowler Cloud API will be used
"prowler-mcp"
]
}
}
}
```
For developers looking to extend the MCP server with new tools or features:
#### HTTP Mode Configuration (Remote Server)
- **[Developer Guide](https://docs.prowler.com/developer-guide/mcp-server)**: Step-by-step instructions for adding new tools
- **[AGENTS.md](./AGENTS.md)**: AI agent guidelines and coding patterns
For HTTP mode, you can configure your MCP client to connect to a remote Prowler MCP server.
## Related Products
Most MCP clients don't natively support HTTP transport with Bearer token authentication. However, you can use the `mcp-remote` proxy tool to connect any MCP client to remote HTTP servers.
##### Using mcp-remote Proxy (Recommended for Claude Desktop)
For clients like Claude Desktop that don't support HTTP transport natively, use the `mcp-remote` npm package as a proxy:
**Using API Key (Recommended):**
```json
{
"mcpServers": {
"prowler": {
"command": "npx",
"args": [
"mcp-remote",
"https://mcp.prowler.com/mcp",
"--header",
"Authorization: Bearer pk_your_api_key_here"
]
}
}
}
```
**Using JWT Token:**
```json
{
"mcpServers": {
"prowler": {
"command": "npx",
"args": [
"mcp-remote",
"https://mcp.prowler.com/mcp",
"--header",
"Authorization: Bearer <your-jwt-token-here>"
]
}
}
}
```
> **Note:** Replace `https://mcp.prowler.com/mcp` with your actual MCP server URL (use `http://localhost:8000/mcp` for local deployment). The `mcp-remote` package is automatically installed by `npx` on first use.
> **Info:** The `mcp-remote` tool acts as a bridge, converting STDIO protocol (used by Claude Desktop) to HTTP requests (used by the remote MCP server). Learn more at [mcp-remote on npm](https://www.npmjs.com/package/mcp-remote).
##### Direct HTTP Configuration (For Compatible Clients)
For clients that natively support HTTP transport with Bearer token authentication:
**Using API Key:**
```json
{
"mcpServers": {
"prowler": {
"url": "https://mcp.prowler.com/mcp",
"headers": {
"Authorization": "Bearer pk_your_api_key_here"
}
}
}
}
```
**Using JWT Token:**
```json
{
"mcpServers": {
"prowler": {
"url": "https://mcp.prowler.com/mcp",
"headers": {
"Authorization": "Bearer <your-jwt-token-here>"
}
}
}
}
```
> **Note:** Replace `mcp.prowler.com` with your actual server hostname and adjust the port if needed (e.g., `http://localhost:8000/mcp` for local deployment).
### Claude Desktop (macOS/Windows)
Add the example server to Claude Desktop's config file, then restart the app.
- macOS: `~/Library/Application Support/Claude/claude_desktop_config.json`
- Windows: `%AppData%\Claude\claude_desktop_config.json` (e.g. `C:\\Users\\<you>\\AppData\\Roaming\\Claude\\claude_desktop_config.json`)
### Cursor (macOS/Linux)
If you want to have it globally available, add the example server to Cursor's config file, then restart the app.
- macOS/Linux: `~/.cursor/mcp.json`
If you want to have it only for the current project, add the example server to the project's root in a new `.cursor/mcp.json` file.
## Documentation
For detailed documentation about the Prowler MCP Server, including guides, tutorials, and use cases, visit the [official Prowler documentation](https://docs.prowler.com).
- **[Prowler Hub](https://hub.prowler.com)**: Browse security checks and compliance frameworks
- **[Prowler Cloud](https://cloud.prowler.com)**: Managed Prowler platform
- **[Lighthouse AI](https://docs.prowler.com/getting-started/products/prowler-lighthouse-ai)**: AI security analyst
## License