### Start Prefect MCP Server with Docker Compose

Source: https://github.com/allen-munsch/mcp-prefect/blob/main/README.md

Use this command to quickly start the Prefect MCP Server using Docker Compose. Ensure Docker is installed and running.

```bash
docker compose up
```

--------------------------------

### Install MCP Prefect

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Install the MCP Prefect package from PyPI or build from source.

```bash
pip install mcp-prefect
```

```bash
git clone https://github.com/allen-munsch/mcp-prefect
cd mcp-prefect
pip install -e .
```

--------------------------------

### Build and Install MCP Prefect from Source

Source: https://github.com/allen-munsch/mcp-prefect/blob/main/README.md

Clone the repository, navigate into the directory, install the package in editable mode, and run the mcp_prefect module.

```bash
git clone https://github.com/allen-munsch/mcp-prefect
cd mcp-prefect
pip install -e .
python -m mcp_prefect
```

--------------------------------

### Install mcp-prefect from Source

Source: https://github.com/allen-munsch/mcp-prefect/blob/main/README.md

Install the Prefect MCP Server directly from its GitHub repository. This is useful for development or if you need the latest unreleased changes.

```bash
git clone https://github.com/allen-munsch/mcp-prefect
cd mcp-prefect
pip install -e .
```

--------------------------------

### Install Dev Dependencies

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Installs development dependencies using uv pip. Ensure you have uv installed.

```bash
uv pip install -e ".[dev"]
```

--------------------------------

### Start Prefect Server

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Starts the Prefect server in detached mode using Docker Compose.

```bash
docker compose up -d
```

--------------------------------

### Install mcp-prefect via pip

Source: https://github.com/allen-munsch/mcp-prefect/blob/main/README.md

Install the Prefect MCP Server package using pip. This is the recommended method for most users.

```bash
pip install mcp-prefect
```

--------------------------------

### Install Dependencies and Run Tests

Source: https://github.com/allen-munsch/mcp-prefect/blob/main/tests/README.md

Installs project dependencies in editable mode with development extras and then runs pytest tests. Ensure Docker is running for any containerized services.

```bash
uv pip install -e ".[dev]"
docker compose up
pytest -svvl tests/
```

--------------------------------

### Run MCP Prefect Server

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Start the MCP Prefect server using Docker Compose or manually with environment variables.

```bash
docker compose up
```

```bash
PREFECT_API_URL=http://localhost:4200/api \
PREFECT_API_KEY=your_api_key_here \
MCP_PORT=8000 \
python -m mcp_prefect.main --transport http
```

--------------------------------

### Get Deployment Details

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Retrieves detailed information about a specific deployment. Use this to inspect schedule configuration, parameters, and other settings.

```python
# MCP Tool Call - Get deployment details
await get_deployment(deployment_id="deploy-123-xyz")
# Response:
# {
#   "id": "deploy-123-xyz",
#   "name": "daily-etl",
#   "flow_id": "flow-abc...",
#   "parameters": {"batch_size": 1000},
#   "schedule": {"cron": "0 6 * * *", "timezone": "UTC"},
#   "is_schedule_active": true,
#   "work_queue_name": "default",
#   "ui_url": "http://localhost:4200/deployments/deploy-123-xyz"
# }
```

--------------------------------

### Get Deployment

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Retrieves detailed information about a specific deployment including schedule configuration and parameters.

```APIDOC
## get_deployment

### Description
Retrieves detailed information about a specific deployment including schedule configuration and parameters.

### Method
GET (assumed, based on action)

### Endpoint
/deployments/{deployment_id}

### Parameters
#### Path Parameters
- **deployment_id** (string) - Required - The ID of the deployment to retrieve.

### Response
#### Success Response (200)
- **id** (string) - The deployment ID.
- **name** (string) - The deployment name.
- **flow_id** (string) - The ID of the associated flow.
- **parameters** (object) - The default parameters for the deployment.
- **schedule** (object) - The schedule configuration.
  - **cron** (string) - The cron expression if applicable.
  - **timezone** (string) - The timezone for the schedule.
- **is_schedule_active** (boolean) - Whether the schedule is active.
- **work_queue_name** (string) - The name of the work queue.
- **ui_url** (string) - The URL to the deployment in the UI.

### Request Example
```json
{
  "deployment_id": "deploy-123-xyz"
}
```

### Response Example
```json
{
  "id": "deploy-123-xyz",
  "name": "daily-etl",
  "flow_id": "flow-abc...",
  "parameters": {"batch_size": 1000},
  "schedule": {"cron": "0 6 * * *", "timezone": "UTC"},
  "is_schedule_active": true,
  "work_queue_name": "default",
  "ui_url": "http://localhost:4200/deployments/deploy-123-xyz"
}
```
```

--------------------------------

### Get Block Types

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Lists available block types in the Prefect instance with optional limit.

```APIDOC
## GET /block-types

### Description
Lists available block types in the Prefect instance.

### Method
GET

### Endpoint
/block-types

### Query Parameters
- **limit** (integer) - Optional - The maximum number of block types to return.

### Response
#### Success Response (200)
- **block_types** (array) - A list of block type objects.
  - **id** (string) - The unique identifier of the block type.
  - **name** (string) - The name of the block type.
  - **slug** (string) - The slug identifier for the block type.

### Response Example
```json
{
  "block_types": [
    {"id": "bt-123...", "name": "Slack Webhook", "slug": "slack-webhook"},
    {"id": "bt-456...", "name": "S3 Bucket", "slug": "s3-bucket"}
  ]
}
```
```

--------------------------------

### Get Variable by Name

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Retrieves a specific variable by its name. Use this to get the current value and metadata of a particular variable.

```python
# MCP Tool Call - Get variable by name
await get_variable(name="api_timeout")
# Response:
# {
#   "id": "var-123...",
#   "name": "api_timeout",
#   "value": 300,
#   "tags": ["config"],
#   "created": "2024-01-15T10:00:00Z",
#   "updated": "2024-01-18T14:30:00Z"
# }
```

--------------------------------

### Get Variables

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Lists variables with optional filtering by name pattern.

```APIDOC
## get_variables

### Description
Lists variables with optional filtering by name pattern.

### Method
GET (assumed, based on action)

### Endpoint
/variables

### Parameters
#### Query Parameters
- **name** (string) - Optional - Filter variables by name (supports pattern matching).
- **limit** (integer) - Optional - The maximum number of variables to return.

### Response
#### Success Response (200)
- **variables** (array) - A list of variable objects.
  - **id** (string) - The variable ID.
  - **name** (string) - The variable name.
  - **value** (any) - The variable value (JSON-serializable).
  - **tags** (array of strings) - Tags associated with the variable.
  - **created** (string) - Timestamp of creation (ISO 8601 format).

### Request Example
```json
{
  "limit": 20
}
```

### Request Example (with name filter)
```json
{
  "name": "config",
  "limit": 10
}
```

### Response Example
```json
{
  "variables": [
    {
      "id": "var-123...",
      "name": "api_timeout",
      "value": 300,
      "tags": ["config"],
      "created": "2024-01-15T10:00:00Z"
    }
  ]
}
```
```

--------------------------------

### Get Automations (Prefect Cloud Only)

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Lists automations with filtering by name and enabled status. This function is specific to Prefect Cloud.

```python
# MCP Tool Call - Get automations
await get_automations(name="*", enabled=True)
```

--------------------------------

### Prefect MCP Server Initialization Log

Source: https://github.com/allen-munsch/mcp-prefect/blob/main/README.md

Log message indicating the successful start of the MCP server with its name and transport protocol. It also confirms successful initialization.

```text
[11/11/25 02:08:06] INFO     Starting MCP server 'MCP Prefect 3.6.1' with transport 'stdio'                                                                                     server.py:1495
✅ Initialized successfully
Server: MCP Prefect 3.6.1 1.14.1
```

--------------------------------

### Get Artifacts

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Lists artifacts with filtering by flow run, task run, type, or key.

```APIDOC
## GET /artifacts

### Description
Lists artifacts with filtering by flow run, task run, type, or key.

### Method
GET

### Endpoint
/artifacts

### Query Parameters
- **flow_run_id** (string) - Optional - Filter artifacts by flow run ID.
- **task_run_id** (string) - Optional - Filter artifacts by task run ID.
- **artifact_type** (string) - Optional - Filter artifacts by type (e.g., "table", "markdown").
- **key** (string) - Optional - Filter artifacts by key.
- **limit** (integer) - Optional - The maximum number of artifacts to return.

### Response
#### Success Response (200)
- **artifacts** (array) - A list of artifact objects.
  - **id** (string) - The unique identifier for the artifact.
  - **key** (string) - The key of the artifact.
  - **type** (string) - The type of the artifact (e.g., "table", "markdown").
  - **data** (object/string) - The artifact data.
  - **flow_run_id** (string) - The ID of the flow run the artifact belongs to.
  - **task_run_id** (string) - The ID of the task run the artifact belongs to.

### Response Example
```json
{
  "artifacts": [
    {
      "id": "artifact-123...",
      "key": "data-summary",
      "type": "table",
      "data": {"columns": ["metric", "value"], "rows": [...]},
      "flow_run_id": "run-123-abc"
    }
  ]
}
```
```

--------------------------------

### Get Enabled Automations

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Retrieves a list of enabled automations with optional limit.

```APIDOC
## GET /automations

### Description
Retrieves a list of enabled automations.

### Method
GET

### Endpoint
/automations

### Query Parameters
- **enabled** (boolean) - Optional - Filter for enabled automations.
- **limit** (integer) - Optional - The maximum number of automations to return.

### Response
#### Success Response (200)
- **automations** (array) - A list of automation objects.
  - **id** (string) - The unique identifier of the automation.
  - **name** (string) - The name of the automation.
  - **enabled** (boolean) - Whether the automation is enabled.
  - **trigger** (object) - The trigger configuration for the automation.
  - **actions** (array) - A list of actions to perform when the trigger fires.

### Response Example
```json
{
  "automations": [
    {
      "id": "auto-123...",
      "name": "Notify on Failure",
      "enabled": true,
      "trigger": {"type": "flow_run_state_change", "state": "FAILED"},
      "actions": [{"type": "send_notification", "channel": "slack"}]
    }
  ]
}
```
```

--------------------------------

### Get Variable

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Retrieves a specific variable by its name.

```APIDOC
## get_variable

### Description
Retrieves a specific variable by its name.

### Method
GET (assumed, based on action)

### Endpoint
/variables/{name}

### Parameters
#### Path Parameters
- **name** (string) - Required - The name of the variable to retrieve.

### Response
#### Success Response (200)
- **id** (string) - The variable ID.
- **name** (string) - The variable name.
- **value** (any) - The variable value (JSON-serializable).
- **tags** (array of strings) - Tags associated with the variable.
- **created** (string) - Timestamp of creation (ISO 8601 format).
- **updated** (string) - Timestamp of last update (ISO 8601 format).

### Request Example
```json
{
  "name": "api_timeout"
}
```

### Response Example
```json
{
  "id": "var-123...",
  "name": "api_timeout",
  "value": 300,
  "tags": ["config"],
  "created": "2024-01-15T10:00:00Z",
  "updated": "2024-01-18T14:30:00Z"
}
```
```

--------------------------------

### Get Deployments

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Lists deployments with filtering by flow name, deployment name, tags, schedule status, and work queue.

```APIDOC
## get_deployments

### Description
Lists deployments with filtering by flow name, deployment name, tags, schedule status, and work queue.

### Method
GET (assumed, based on action)

### Endpoint
/deployments

### Parameters
#### Query Parameters
- **flow_name** (string) - Optional - Filter by flow name.
- **deployment_name** (string) - Optional - Filter by deployment name.
- **tags** (array of strings) - Optional - Filter by tags.
- **is_schedule_active** (boolean) - Optional - Filter by schedule active status.
- **work_queue_name** (string) - Optional - Filter by work queue name.
- **limit** (integer) - Optional - The maximum number of deployments to return.

### Response
#### Success Response (200)
- **deployments** (array) - A list of deployment objects.
  - **id** (string) - The deployment ID.
  - **name** (string) - The deployment name.
  - **flow_id** (string) - The ID of the associated flow.
  - **is_schedule_active** (boolean) - Whether the schedule is active.
  - **schedule** (object) - The schedule configuration.
    - **cron** (string) - The cron expression if applicable.
    - **timezone** (string) - The timezone for the schedule.
  - **ui_url** (string) - The URL to the deployment in the UI.

### Request Example
```json
{
  "is_schedule_active": true,
  "limit": 10
}
```

### Request Example (with tags)
```json
{
  "flow_name": "data-processing",
  "tags": ["production"]
}
```

### Response Example
```json
{
  "deployments": [
    {
      "id": "deploy-123...",
      "name": "daily-etl",
      "flow_id": "flow-abc...",
      "is_schedule_active": true,
      "schedule": {"cron": "0 6 * * *"},
      "ui_url": "http://localhost:4200/deployments/deploy-123..."
    }
  ]
}
```
```

--------------------------------

### Get Deployments

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Lists deployments with filtering capabilities. Useful for finding active deployments or deployments matching specific criteria like flow name or tags.

```python
# MCP Tool Call - Get active deployments
await get_deployments(is_schedule_active=True, limit=10)
# Response:
# {
#   "deployments": [
#     {
#       "id": "deploy-123...",
#       "name": "daily-etl",
#       "flow_id": "flow-abc...",
#       "is_schedule_active": true,
#       "schedule": {"cron": "0 6 * * *"},
#       "ui_url": "http://localhost:4200/deployments/deploy-123..."
#     }
#   ]
# }
```

```python
# Get deployments by flow name
await get_deployments(flow_name="data-processing", tags=["production"])
```

--------------------------------

### Get Prefect Flow Runs

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

List Prefect flow runs with filtering by state, deployment, tags, and time ranges. The response includes UI URLs.

```python
# MCP Tool Call - Get running flow runs
await get_flow_runs(state_type="RUNNING", limit=10)
```

```python
# Get failed flow runs for a specific deployment
await get_flow_runs(
    deployment_id="deploy-456...",
    state_type="FAILED",
    start_time_after="2024-01-15T00:00:00Z",
    limit=20
)
```

```python
# Get flow runs by flow name
await get_flow_runs(flow_name="etl-pipeline", limit=5)
```

--------------------------------

### Get Variables

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Lists variables with optional filtering by name pattern. Useful for retrieving configuration values or other stored data.

```python
# MCP Tool Call - Get all variables
await get_variables(limit=20)
# Response:
# {
#   "variables": [
#     {
#       "id": "var-123...",
#       "name": "api_timeout",
#       "value": 300,
#       "tags": ["config"],
#       "created": "2024-01-15T10:00:00Z"
#     }
#   ]
# }
```

```python
# Filter variables by name
await get_variables(name="config", limit=10)
```

--------------------------------

### Get Automations (Prefect Cloud Only)

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Lists automations with filtering by name and enabled status. This endpoint is only available in Prefect Cloud.

```APIDOC
## GET /automations

### Description
Lists automations with filtering by name and enabled status. This endpoint is only available in Prefect Cloud.

### Method
GET

### Endpoint
/automations

### Query Parameters
- **name** (string) - Optional - Filter automations by name.
- **is_enabled** (boolean) - Optional - Filter automations by enabled status.
- **limit** (integer) - Optional - The maximum number of automations to return.

### Response
#### Success Response (200)
- **automations** (array) - A list of automation objects.
  - **id** (string) - The unique identifier for the automation.
  - **name** (string) - The name of the automation.
  - **is_enabled** (boolean) - Indicates if the automation is enabled.
  - **description** (string) - Description of the automation.

### Response Example
```json
{
  "automations": [
    {
      "id": "auto-123...",
      "name": "High CPU Alert",
      "is_enabled": true,
      "description": "Alerts when CPU usage exceeds 90%"
    }
  ]
}
```
```

--------------------------------

### Get Work Queues

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Lists work queues with filtering by name, paused status, and work pool.

```APIDOC
## GET /work_queues

### Description
Lists work queues with filtering by name, paused status, and work pool.

### Method
GET

### Endpoint
/work_queues

### Query Parameters
- **name** (string) - Optional - Filter work queues by name.
- **is_paused** (boolean) - Optional - Filter work queues by paused status.
- **work_pool_name** (string) - Optional - Filter work queues by work pool name.
- **limit** (integer) - Optional - The maximum number of work queues to return.

### Response
#### Success Response (200)
- **work_queues** (array) - A list of work queue objects.
  - **id** (string) - The unique identifier for the work queue.
  - **name** (string) - The name of the work queue.
  - **is_paused** (boolean) - Indicates if the work queue is paused.
  - **concurrency_limit** (integer) - The concurrency limit for the work queue.
  - **work_pool_name** (string) - The name of the work pool the queue belongs to.

### Response Example
```json
{
  "work_queues": [
    {
      "id": "queue-123...",
      "name": "default",
      "is_paused": false,
      "concurrency_limit": 10,
      "work_pool_name": "default-pool"
    }
  ]
}
```
```

--------------------------------

### Get Block Documents

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Retrieves configured block instances by block type slug. Use to find specific block configurations, like Slack webhooks.

```python
await get_block_documents(
    block_type_slug="slack-webhook",
    limit=10
)
```

--------------------------------

### Get Task Runs

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Lists task runs with filtering by task name, state, tags, and time ranges.

```APIDOC
## get_task_runs

### Description
Lists task runs with filtering by task name, state, tags, and time ranges.

### Method
GET (assumed, based on action)

### Endpoint
/task_runs

### Parameters
#### Query Parameters
- **task_name** (string) - Optional - Filter by task name.
- **state_type** (string) - Optional - Filter by state type (e.g., FAILED, RUNNING, COMPLETED).
- **tags** (array of strings) - Optional - Filter by tags.
- **start_time_min** (string) - Optional - Filter by minimum start time (ISO 8601 format).
- **start_time_max** (string) - Optional - Filter by maximum start time (ISO 8601 format).
- **end_time_min** (string) - Optional - Filter by minimum end time (ISO 8601 format).
- **end_time_max** (string) - Optional - Filter by maximum end time (ISO 8601 format).
- **limit** (integer) - Optional - The maximum number of task runs to return.

### Response
#### Success Response (200)
- **task_runs** (array) - A list of task run objects.
  - **id** (string) - The task run ID.
  - **name** (string) - The task name.
  - **state** (object) - The state of the task run.
    - **type** (string) - The type of the state.
  - **tags** (array of strings) - Tags associated with the task run.
  - **start_time** (string) - Timestamp of when the task run started (ISO 8601 format).
  - **end_time** (string) - Timestamp of when the task run ended (ISO 8601 format).

### Request Example
```json
{
  "state_type": "FAILED",
  "limit": 10
}
```

### Response Example
```json
{
  "task_runs": [
    {
      "id": "task-run-123...",
      "name": "process_data",
      "state": {"type": "FAILED"},
      "tags": ["data-quality"],
      "start_time": "2024-01-18T10:00:00Z",
      "end_time": "2024-01-18T10:05:00Z"
    }
  ]
}
```
```

--------------------------------

### Get Enabled Automations

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Retrieves a list of enabled automations with optional limit. Useful for monitoring active automation rules.

```python
await get_automations(enabled=True, limit=10)
```

--------------------------------

### Get Block Documents

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Retrieves block documents (configured block instances) by type with optional limit.

```APIDOC
## GET /block-documents

### Description
Retrieves block documents (configured block instances) by type.

### Method
GET

### Endpoint
/block-documents

### Query Parameters
- **block_type_slug** (string) - Required - The slug of the block type to filter by.
- **limit** (integer) - Optional - The maximum number of block documents to return.

### Response
#### Success Response (200)
- **block_documents** (array) - A list of block document objects.
  - **id** (string) - The unique identifier of the block document.
  - **name** (string) - The name of the block document.
  - **block_type_slug** (string) - The slug of the block type.

### Response Example
```json
{
  "block_documents": [
    {
      "id": "block-123...",
      "name": "team-notifications",
      "block_type_slug": "slack-webhook"
    }
  ]
}
```
```

--------------------------------

### Get Specific Prefect Flow

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Retrieve detailed information for a specific Prefect flow using its UUID. The response includes the UI URL.

```python
# MCP Tool Call - Get specific flow details
await get_flow(flow_id="abc123-def456-ghi789")
```

--------------------------------

### Get Prefect Flows

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Retrieve a list of Prefect flows, with options to filter by name, tags, and creation date. The response includes UI URLs for direct access.

```python
# MCP Tool Call - Get all flows
await get_flows(limit=10)
```

```python
# Get flows by name pattern
await get_flows(flow_name="etl", tags=["production"], limit=5)
```

```python
# Get flows created after a specific date
await get_flows(created_after="2024-01-01T00:00:00Z", limit=20)
```

--------------------------------

### Get Specific Prefect Flow Run

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Retrieve detailed information about a specific Prefect flow run, including its state, timing, and parameters. The response includes the UI URL.

```python
# MCP Tool Call - Get flow run details
await get_flow_run(flow_run_id="run-123-abc-def")
```

--------------------------------

### Get Artifacts

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Lists artifacts with filtering by flow run, task run, type, or key. Useful for retrieving stored data, tables, or markdown content.

```python
# MCP Tool Call - Get artifacts for a flow run
await get_artifacts(
    flow_run_id="run-123-abc",
    artifact_type="table",
    limit=10
)
# Response:
# {
#   "artifacts": [
#     {
#       "id": "artifact-123...",
#       "key": "data-summary",
#       "type": "table",
#       "data": {"columns": ["metric", "value"], "rows": [...]},
#       "flow_run_id": "run-123-abc"
#     }
#   ]
# }
```

--------------------------------

### Get Task Runs

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Lists task runs with filtering by task name, state, tags, and time ranges. Useful for monitoring task execution and debugging.

```python
# MCP Tool Call - Get failed task runs
await get_task_runs(state_type="FAILED", limit=10)
# Response:
# {
#   "task_runs": [
#     {

```

--------------------------------

### Get Task Run Logs

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Retrieves logs for a specific task run, with optional filtering by log level. Useful for debugging individual task executions.

```python
# MCP Tool Call - Get task run logs
await get_task_run_logs(
    task_run_id="task-123-abc",
    limit=50,
    level=30  # WARNING level and above
)
# Response:
# {
#   "task_run_logs": [
#     {"timestamp": "2024-01-20T14:05:00Z", "level": 30, "message": "Slow response from API", "name": "prefect.task"}
#   ]
# }
```

--------------------------------

### Running Tests

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Instructions for setting up the development environment and running tests for MCP Prefect.

```APIDOC
## Running Tests

### Description
Instructions to install dependencies, start the Prefect server, and run tests.

### Commands
1. **Install dev dependencies**
   ```bash
   uv pip install -e ".[dev]"
   ```

2. **Start Prefect server**
   ```bash
   docker compose up -d
   ```

3. **Run tests**
   ```bash
   pytest -svvl tests/
   ```
```

--------------------------------

### Prefect MCP Server Tool Listing

Source: https://github.com/allen-munsch/mcp-prefect/blob/main/README.md

Output showing the total number of tools available and a categorized list of tools supported by the Prefect MCP Server, including Artifact, Automation, Block, Deployment, Flow, Log, Other, Task, Variable, and Work.

```text
🔄 Listing tools...

🎯 FOUND 64 TOOLS:
================================================================================

📂 ARTIFACT (6 tools)
  🔧 create_artifact
  🔧 delete_artifact
  🔧 get_artifact
  🔧 get_artifacts
  🔧 get_latest_artifacts
  🔧 update_artifact

📂 AUTOMATION (7 tools)
  🔧 create_automation
  🔧 delete_automation
  🔧 get_automation
  🔧 get_automations
  🔧 pause_automation
  🔧 resume_automation
  🔧 update_automation

📂 BLOCK (5 tools)
  🔧 delete_block_document
  🔧 get_block_document
  🔧 get_block_documents
  🔧 get_block_type
  🔧 get_block_types

📂 DEPLOYMENT (8 tools)
  🔧 delete_deployment
  🔧 get_deployment
  🔧 get_deployment_schedule
  🔧 get_deployments
  🔧 pause_deployment_schedule
  🔧 resume_deployment_schedule
  🔧 set_deployment_schedule
  🔧 update_deployment

📂 FLOW (13 tools)
  🔧 cancel_flow_run
  🔧 create_flow_run_from_deployment
  🔧 delete_flow
  🔧 delete_flow_run
  🔧 get_flow
  🔧 get_flow_run
  🔧 get_flow_run_logs
  🔧 get_flow_runs
  🔧 get_flow_runs_by_flow
  🔧 get_flows
  🔧 get_task_runs_by_flow_run
  🔧 restart_flow_run
  🔧 set_flow_run_state

📂 LOG (2 tools)
  🔧 create_log
  🔧 get_logs

📂 OTHER (1 tools)
  🔧 get_health

📂 TASK (4 tools)
  🔧 get_task_run
  🔧 get_task_run_logs
  🔧 get_task_runs
  🔧 set_task_run_state

📂 VARIABLE (5 tools)
  🔧 create_variable
  🔧 delete_variable
  🔧 get_variable
  🔧 get_variables
  🔧 update_variable

📂 WORK (13 tools)
  🔧 create_work_queue
  🔧 delete_work_queue
  🔧 get_current_workspace
  🔧 get_work_queue
  🔧 get_work_queue_by_name
  🔧 get_work_queue_runs
  🔧 get_work_queues
  🔧 get_workspace
  🔧 get_workspace_by_handle
  🔧 get_workspaces
  🔧 pause_work_queue
  🔧 resume_work_queue
  🔧 update_work_queue

📊 TOTAL: 64 tools across 10 categories
```

--------------------------------

### Get Health Status

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Checks the health status of the connected Prefect server.

```APIDOC
## GET /health

### Description
Checks the health status of the connected Prefect server.

### Method
GET

### Endpoint
/health

### Response
#### Success Response (200)
- **status** (string) - The health status of the server (e.g., "OK").

### Response Example
```
"OK"
```
```

--------------------------------

### Platform Integration - Cursor / Windsurf

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Configuration for integrating MCP Prefect into IDEs like Cursor and Windsurf via MCP.

```APIDOC
## MCP Configuration for Cursor / Windsurf

### Description
Add MCP Prefect to your IDE's MCP configuration.

### Configuration
```json
{
  "mcpServers": {
    "prefect": {
      "command": "mcp-prefect",
      "args": ["--transport", "stdio"],
      "env": {
        "PREFECT_API_URL": "http://localhost:4200/api",
        "PREFECT_API_KEY": "your_api_key_here"
      }
    }
  }
}
```
```

--------------------------------

### Create Work Queue

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Creates a new work queue with optional configuration.

```APIDOC
## POST /work_queues

### Description
Creates a new work queue with optional configuration.

### Method
POST

### Endpoint
/work_queues

### Request Body
- **name** (string) - Required - The name of the new work queue.
- **description** (string) - Optional - A description for the work queue.
- **concurrency_limit** (integer) - Optional - The concurrency limit for the work queue.
- **work_pool_name** (string) - Optional - The name of the work pool to associate with the queue.

### Request Example
```json
{
  "name": "high-priority",
  "description": "Queue for high-priority jobs",
  "concurrency_limit": 5,
  "work_pool_name": "production-pool"
}
```

### Response
#### Success Response (201)
- **id** (string) - The unique identifier for the newly created work queue.
- **name** (string) - The name of the work queue.
- **description** (string) - The description of the work queue.
- **concurrency_limit** (integer) - The concurrency limit for the work queue.
- **is_paused** (boolean) - Indicates if the work queue is paused (defaults to false).

### Response Example
```json
{
  "id": "queue-new-123...",
  "name": "high-priority",
  "description": "Queue for high-priority jobs",
  "concurrency_limit": 5,
  "is_paused": false
}
```
```

--------------------------------

### Configure Gemini CLI for MCP Prefect

Source: https://github.com/allen-munsch/mcp-prefect/blob/main/README.md

Use this command to set up the Gemini CLI for MCP Prefect integration, specifying the command and transport protocol.

```bash
gemini config set mcp-servers.prefect "mcp-prefect --transport stdio"
```

--------------------------------

### Create Prefect Flow Run from Deployment

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Trigger a new Prefect flow run from an existing deployment. Supports specifying parameters, name, and tags.

```python
# MCP Tool Call - Create flow run from deployment
await create_flow_run_from_deployment(
    deployment_id="deploy-456-xyz",
    parameters={"batch_size": 500, "environment": "staging"},
    name="manual-trigger-20240120",
    tags=["manual", "test"]
)
```

```python
# Simple trigger without parameters
await create_flow_run_from_deployment(deployment_id="deploy-456-xyz")
```

--------------------------------

### Get Task Run Logs

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Retrieves logs for a specific task run with optional level filtering.

```APIDOC
## GET /task_runs/{task_run_id}/logs

### Description
Retrieves logs for a specific task run with optional level filtering.

### Method
GET

### Endpoint
/task_runs/{task_run_id}/logs

### Path Parameters
- **task_run_id** (string) - Required - The ID of the task run.

### Query Parameters
- **limit** (integer) - Optional - The maximum number of log entries to return.
- **level** (integer) - Optional - Filter logs by level (e.g., 30 for WARNING and above).

### Response
#### Success Response (200)
- **task_run_logs** (array) - A list of log entries.
  - **timestamp** (string) - The timestamp of the log entry.
  - **level** (integer) - The log level.
  - **message** (string) - The log message.
  - **name** (string) - The name associated with the log entry.

### Response Example
```json
{
  "task_run_logs": [
    {"timestamp": "2024-01-20T14:05:00Z", "level": 30, "message": "Slow response from API", "name": "prefect.task"}
  ]
}
```
```

--------------------------------

### Configure Claude Desktop for MCP Prefect

Source: https://github.com/allen-munsch/mcp-prefect/blob/main/README.md

Add this configuration to your 'claude_desktop_config.json' to enable MCP Prefect integration. It specifies the command and transport protocol.

```json
{
  "mcpServers": {
    "prefect": {
      "command": "mcp-prefect",
      "args": ["--transport", "stdio"]
    }
  }
}
```

--------------------------------

### Get Flow Run Logs

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Retrieves logs for a specific flow run with optional level filtering.

```APIDOC
## GET /flow_runs/{flow_run_id}/logs

### Description
Retrieves logs for a specific flow run with optional level filtering.

### Method
GET

### Endpoint
/flow_runs/{flow_run_id}/logs

### Path Parameters
- **flow_run_id** (string) - Required - The ID of the flow run.

### Query Parameters
- **limit** (integer) - Optional - The maximum number of log entries to return.
- **level** (integer) - Optional - Filter logs by level (e.g., 20 for INFO and above).

### Response
#### Success Response (200)
- **flow_run_logs** (array) - A list of log entries.
  - **timestamp** (string) - The timestamp of the log entry.
  - **level** (integer) - The log level.
  - **message** (string) - The log message.
  - **name** (string) - The name associated with the log entry.

### Response Example
```json
{
  "flow_run_logs": [
    {"timestamp": "2024-01-20T14:00:00Z", "level": 20, "message": "Starting flow...", "name": "prefect.flow"},
    {"timestamp": "2024-01-20T14:00:01Z", "level": 20, "message": "Processing batch 1...", "name": "prefect.task"},
    {"timestamp": "2024-01-20T14:15:00Z", "level": 20, "message": "Flow completed successfully", "name": "prefect.flow"}
  ]
}
```
```

--------------------------------

### Platform Integration - Claude Desktop

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Configuration for integrating MCP Prefect with Claude Desktop for AI-powered workflow management.

```APIDOC
## MCP Configuration for Claude Desktop

### Description
Configure MCP Prefect within Claude Desktop.

### Configuration
```json
{
  "mcpServers": {
    "prefect": {
      "command": "mcp-prefect",
      "args": ["--transport", "stdio"],
      "env": {
        "PREFECT_API_URL": "http://localhost:4200/api",
        "PREFECT_API_KEY": "your_api_key_here"
      }
    }
  }
}
```
```

--------------------------------

### Create Work Queue

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Creates a new work queue with optional configuration such as name, description, concurrency limit, and work pool. Useful for setting up dedicated queues for specific job types.

```python
# MCP Tool Call - Create work queue
await create_work_queue(
    name="high-priority",
    description="Queue for high-priority jobs",
    concurrency_limit=5,
    work_pool_name="production-pool"
)
# Response:
# {
#   "id": "queue-new-123...",
#   "name": "high-priority",
#   "description": "Queue for high-priority jobs",
#   "concurrency_limit": 5,
#   "is_paused": false
# }
```

--------------------------------

### Create Automation

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Creates a new automation with specified triggers, actions, and description. Use for setting up event-driven workflow responses.

```python
await create_automation(
    name="Alert on ETL Failure",
    enabled=True,
    trigger={
        "type": "flow_run_state_change",
        "match": {"prefect.resource.name": "daily-etl"},
        "to_state": {"type": "FAILED"}
    },
    actions=[
        {"type": "send_notification", "block_document_id": "slack-block-id"}
    ],
    description="Alerts the team when daily ETL fails"
)
```

--------------------------------

### Manually Run Prefect MCP Server

Source: https://github.com/allen-munsch/mcp-prefect/blob/main/README.md

Run the Prefect MCP Server manually from the command line. You need to set environment variables for Prefect API URL, API Key, and the MCP port.

```bash
PREFECT_API_URL=http://localhost:4200/api \
PREFECT_API_KEY=your_api_key_here \
MCP_PORT=8000 \
python -m mcp_prefect.main --transport http
```

--------------------------------

### Create Automation

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Creates a new automation with specified triggers and actions. This feature is available for Prefect Cloud only.

```APIDOC
## POST /automations

### Description
Creates a new automation with triggers and actions.

### Method
POST

### Endpoint
/automations

### Request Body
- **name** (string) - Required - The name of the automation.
- **enabled** (boolean) - Required - Whether the automation is enabled.
- **trigger** (object) - Required - The trigger configuration.
  - **type** (string) - Required - The type of trigger (e.g., "flow_run_state_change").
  - **match** (object) - Optional - Criteria to match for the trigger.
  - **to_state** (object) - Optional - The target state for state change triggers.
- **actions** (array) - Required - A list of actions to perform.
  - **type** (string) - Required - The type of action (e.g., "send_notification").
  - **block_document_id** (string) - Required - The ID of the block document to use for the action.
- **description** (string) - Optional - A description for the automation.

### Request Example
```json
{
  "name": "Alert on ETL Failure",
  "enabled": true,
  "trigger": {
    "type": "flow_run_state_change",
    "match": {"prefect.resource.name": "daily-etl"},
    "to_state": {"type": "FAILED"}
  },
  "actions": [
    {"type": "send_notification", "block_document_id": "slack-block-id"}
  ],
  "description": "Alerts the team when daily ETL fails"
}
```
```

--------------------------------

### Set Deployment Schedule

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Configures a cron or interval-based schedule for a deployment. Use this to define when a deployment should automatically run.

```python
# MCP Tool Call - Set cron schedule
await set_deployment_schedule(
    deployment_id="deploy-123-xyz",
    cron="0 6 * * *",
    timezone="America/New_York"
)
# Response: {"cron": "0 6 * * *", "timezone": "America/New_York"}
```

```python
# Set interval schedule (every 30 minutes)
await set_deployment_schedule(
    deployment_id="deploy-123-xyz",
    interval_seconds=1800,
    anchor_date="2024-01-01T00:00:00Z",
    timezone="UTC"
)
```

--------------------------------

### List Block Types

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Lists available block types in the Prefect instance. Useful for discovering configurable components.

```python
await get_block_types(limit=20)
```

--------------------------------

### Get Task Runs

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Retrieves task runs, with options to filter by name pattern, state type, and limit the number of results.

```APIDOC
## GET /task_runs

### Description
Retrieves task runs, with options to filter by name pattern, state type, and limit the number of results.

### Method
GET

### Endpoint
/task_runs

### Query Parameters
- **task_name** (string) - Optional - Filter task runs by name pattern.
- **state_type** (string) - Optional - Filter task runs by state type (e.g., "COMPLETED", "FAILED").
- **limit** (integer) - Optional - The maximum number of task runs to return.

### Response
#### Success Response (200)
- **task_runs** (array) - A list of task run objects.
  - **id** (string) - The unique identifier for the task run.
  - **name** (string) - The name of the task run.
  - **flow_run_id** (string) - The ID of the flow run this task belongs to.
  - **state_type** (string) - The state type of the task run (e.g., "FAILED", "COMPLETED").
  - **state_name** (string) - The name of the task run state.
  - **ui_url** (string) - The URL to view the task run in the UI.

### Response Example
```json
{
  "task_runs": [
    {
      "id": "task-123...",
      "name": "extract_data",
      "flow_run_id": "run-abc...",
      "state_type": "FAILED",
      "state_name": "Failed",
      "ui_url": "http://localhost:4200/task-runs/task-123..."
    }
  ]
}
```
```

--------------------------------

### Create Log

Source: https://context7.com/allen-munsch/mcp-prefect/llms.txt

Creates a log entry associated with a flow run or task run.

```APIDOC
## POST /logs

### Description
Creates a log entry associated with a flow run or task run.

### Method
POST

### Endpoint
/logs

### Request Body
- **message** (string) - Required - The log message.
- **level** (integer) - Required - The log level (e.g., 30 for WARNING).
- **flow_run_id** (string) - Optional - The ID of the flow run to associate the log with.
- **task_run_id** (string) - Optional - The ID of the task run to associate the log with.
- **name** (string) - Optional - A name for the log entry.

### Request Example
```json
{
  "message": "Manual intervention: Skipped validation step",
  "level": 30,
  "flow_run_id": "run-123-abc",
  "name": "prefect.manual"
}
```

### Response
#### Success Response (201)
- **message** (string) - Confirmation message.

### Response Example
```
"Log created successfully."
```
```