Flint

https://www.tryflint.com/docs
Admin
Flint is an AI-powered visual website builder that enables teams to create, edit, and publish...
Tokens:22,196
Snippets:355
Trust Score:7.9
Update:6 months ago
Show doc for...
Tokens:
Raw
### Install Flint Plugin

Source: https://www.tryflint.com/docs/claude-code-plugin

Use this command within Claude Code to install the Flint plugin from the marketplace.

```text
/plugin install flint
```

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

### Create an Agent Task

Source: https://www.tryflint.com/docs

Start a background task to modify a site using a prompt.

```bash
curl -X POST https://app.tryflint.com/api/v1/agent/tasks \
 -H "Authorization: Bearer ak_your_api_key_here" \
 -H "Content-Type: application/json" \
 -d '{ "siteId": "550e8400-e29b-41d4-a716-446655440000", "prompt": "Add a new About page with a team section" }'
```

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

### Create Prompt Task Request

Source: https://www.tryflint.com/docs/api

Example request for creating a task using free-form AI instructions.

```bash
curl -X POST https://app.tryflint.com/api/v1/agent/tasks \
  -H "Authorization: Bearer ak_your_api_key_here" \
  -H "Content-Type: application/json" \
  -d '{
    "siteId": "550e8400-e29b-41d4-a716-446655440000",
    "prompt": "Add a new About page with a team section",
    "callbackUrl": "https://your-server.com/webhooks/flint"
  }'
```

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

### Claude MCP Integration Setup

Source: https://www.tryflint.com/docs/claude-mcp

Instructions on how to connect Flint to Claude using the MCP server URL.

```APIDOC
## Claude MCP Integration Setup

### Description
Connect Flint to Claude to create landing pages by describing them. This integration uses Flint's MCP (Model Context Protocol) server.

### Server URL
`https://mcp.tryflint.com/mcp`

### Setup Steps
1. Go to `claude.ai/customize/connectors`.
2. Click the plus sign to add a custom connector.
3. Paste the Flint MCP server URL: `https://mcp.tryflint.com/mcp`.
4. Sign in to your Flint account to authenticate.
5. Start a new Claude conversation and describe the page you want to create.
```

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

### Create Generate Pages Task Request

Source: https://www.tryflint.com/docs/api

Example request for creating multiple pages based on a template.

```bash
curl -X POST https://app.tryflint.com/api/v1/agent/tasks \
  -H "Authorization: Bearer ak_your_api_key_here" \
  -H "Content-Type: application/json" \
  -d '{
    "siteId": "550e8400-e29b-41d4-a716-446655440000",
    "command": "generate_pages",
    "templatePageSlug": "/case-studies/template",
    "items": [
      {
        "targetPageSlug": "/case-studies/acme-corp",
        "context": "Company: Acme Corp. Industry: Manufacturing. Challenge: Needed to reduce production downtime. Solution: Implemented our monitoring system. Results: 40% reduction in downtime, $2M annual savings."
      },
      {
        "targetPageSlug": "/case-studies/globex",
        "context": "Company: Globex Inc. Industry: Finance. Challenge: Manual compliance reporting taking weeks. Solution: Automated reporting pipeline. Results: Reports generated in hours, 99.9% accuracy."
      }
    ],
    "callbackUrl": "https://your-server.com/webhooks/flint"
  }'
```

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

### run_background_agent

Source: https://www.tryflint.com/docs/claude-mcp

Starts a background AI agent on a Flint site to modify the website based on a prompt.

```APIDOC
## POST /run_background_agent

### Description
Starts a background AI agent on a Flint site. The agent executes a prompt to modify your website.

### Method
POST

### Endpoint
`/run_background_agent`

### Parameters
#### Request Body
- **siteId** (string (UUID)) - Required - The UUID of the Flint site to modify.
- **prompt** (string) - Required - What the background agent should do.

### Request Example
```json
{
  "siteId": "your-site-uuid",
  "prompt": "Create a new section with a headline and a call to action."
}
```

### Response
#### Success Response (200)
- **workflowId** (string) - The ID that can be used to check the status of the background agent.

#### Response Example
```json
{
  "workflowId": "workflow-abc-123"
}
```
```

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

### Get Task Status Request

Source: https://www.tryflint.com/docs/api

Example request to check the status of a background task.

```bash
curl https://app.tryflint.com/api/v1/agent/tasks/bg-550e8400-a1b2c3d4 \
  -H "Authorization: Bearer ak_your_api_key_here"
```

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

### Task Creation Success Response

Source: https://www.tryflint.com/docs/api

The JSON response returned upon successfully starting a task.

```json
{
  "taskId": "bg-550e8400-e29b-41d4-a716-446655440000-a1b2c3d4",
  "status": "running",
  "createdAt": "2026-03-11T12:00:00.000Z"
}
```

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

### Run Background Agent Tool

Source: https://www.tryflint.com/docs/claude-mcp

Starts a background AI agent on a Flint site to modify the website based on a given prompt. Requires the site ID and the prompt. Returns a workflow ID to track the agent's status.

```json
{
  "tool_code": "run_background_agent",
  "parameters": {
    "siteId": "<site_uuid>",
    "prompt": "<your_prompt>"
  }
}
```

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

### POST /agent/tasks

Source: https://www.tryflint.com/docs

Starts a new background agent task to modify a specific Flint site.

```APIDOC
## POST /agent/tasks

### Description
Start a new background agent task to modify a site.

### Method
POST

### Endpoint
/agent/tasks

### Parameters
#### Request Body
- **siteId** (string) - Required - UUID of the site to modify
- **prompt** (string) - Required - Instructions for the agent to follow

### Request Example
{
  "siteId": "550e8400-e29b-41d4-a716-446655440000",
  "prompt": "Add a new About page with a team section"
}

### Response
#### Success Response (200)
- **taskId** (string) - The unique identifier for the created task

#### Response Example
{
  "taskId": "bg-550e8400-e29b-41d4-a716-446655440000-a1b2c3d4"
}
```

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

### Generate Page Prompt

Source: https://www.tryflint.com/docs/claude-code-mcp

Example of a natural language prompt used to instruct Claude Code to generate a personalized landing page.

```text
Generate a personalized account page targeting Anthropic. Tailor the messaging to address their specific industry needs.
```

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

### POST /agent/tasks - Create a Task

Source: https://www.tryflint.com/docs/api

Starts a new background agent task to modify a site. Supports two task types: 'prompt' for free-form instructions and 'generate_pages' for creating multiple pages from a template.

```APIDOC
## POST /agent/tasks

### Description
Starts a new background agent task to modify a site.

### Method
POST

### Endpoint
/agent/tasks

### Parameters
#### Request Body
**Task Type 1: Prompt**
- **siteId** (string) - Required - UUID of the site to modify
- **prompt** (string) - Required - What you want the AI to do, in plain English
- **command** (string) - Optional - Set to `prompt` or omit entirely (default)
- **callbackUrl** (string) - Optional - HTTPS URL to receive a POST when the task completes

**Task Type 2: Generate Pages**
- **siteId** (string) - Required - UUID of the site to modify
- **command** (string) - Required - Must be `generate_pages`
- **templatePageSlug** (string) - Required - The path of any existing page on your Flint site to use as the design template
- **items** (array) - Required - 1-10 items to generate. Each item has: `targetPageSlug` (required, the path for the new page) and `context` (required, the content or data for this specific page)
- **callbackUrl** (string) - Optional - HTTPS URL to receive a POST when the task completes

### Request Example (Prompt)
```json
{
  "siteId": "550e8400-e29b-41d4-a716-446655440000",
  "prompt": "Add a new About page with a team section",
  "callbackUrl": "https://your-server.com/webhooks/flint"
}
```

### Request Example (Generate Pages)
```json
{
  "siteId": "550e8400-e29b-41d4-a716-446655440000",
  "command": "generate_pages",
  "templatePageSlug": "/case-studies/template",
  "items": [
    {
      "targetPageSlug": "/case-studies/acme-corp",
      "context": "Company: Acme Corp. Industry: Manufacturing. Challenge: Needed to reduce production downtime. Solution: Implemented our monitoring system. Results: 40% reduction in downtime, $2M annual savings."
    },
    {
      "targetPageSlug": "/case-studies/globex",
      "context": "Company: Globex Inc. Industry: Finance. Challenge: Manual compliance reporting taking weeks. Solution: Automated reporting pipeline. Results: Reports generated in hours, 99.9% accuracy."
    }
  ],
  "callbackUrl": "https://your-server.com/webhooks/flint"
}
```

### Response
#### Success Response (200)
- **taskId** (string) - The ID of the created task
- **status** (string) - The initial status of the task (e.g., "running")
- **createdAt** (string) - The timestamp when the task was created

#### Response Example
```json
{
  "taskId": "bg-550e8400-e29b-41d4-a716-446655440000-a1b2c3d4",
  "status": "running",
  "createdAt": "2026-03-11T12:00:00.000Z"
}
```

### Error Handling
- **400**: Site is missing repository information or Invalid callback URL
- **404**: Site not found
- **500**: Failed to start task
```

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

### Example Webhook Handler (Node.js)

Source: https://www.tryflint.com/docs/api

This Node.js Express.js code snippet demonstrates how to handle incoming webhook POST requests from Flint. It checks the task status and logs relevant information or errors. It is crucial to return a 200 status code to acknowledge receipt.

```javascript
app.post('/webhooks/flint', (req, res) => {
  const { taskId, status, pages, error } = req.body;
  if (status === 'succeeded') {
    console.log(`Task ${taskId} completed. Pages:`, pages);
  } else {
    console.error(`Task ${taskId} failed:`, error);
  }
  res.sendStatus(200);
});
```

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

### Prompt to Generate GEO Landing Pages

Source: https://www.tryflint.com/docs/seo-landing-pages

This prompt guides Flint in generating GEO landing pages from a CSV file. It specifies how to incorporate location-specific data, testimonials, and map embeds into each page.

```text
Using this CSV, generate GEO landing pages for each city.
For each page:
- Include [city] in H1, title tag, and meta description
- Add location-specific stats from the data
- Use the city's testimonial if available
- Include a local map embed placeholder
```

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

### Check Background Agent Status Tool

Source: https://www.tryflint.com/docs/claude-mcp

Checks the status of a background agent previously started with `run_background_agent`. Requires the workflow ID returned from the initial call. Returns the current status, phase, and output upon completion.

```json
{
  "tool_code": "check_background_agent_status",
  "parameters": {
    "workflowId": "<workflow_id>"
  }
}
```

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

### GET /agent/tasks/{taskId} - Get Task Status

Source: https://www.tryflint.com/docs/api

Retrieves the current status and results of a previously created agent task.

```APIDOC
## GET /agent/tasks/{taskId}

### Description
Retrieve the current status and results of a task.

### Method
GET

### Endpoint
/agent/tasks/{taskId}

### Parameters
#### Path Parameters
- **taskId** (string) - Required - The task ID from creation

### Request Example
```bash
curl https://app.tryflint.com/api/v1/agent/tasks/bg-550e8400-a1b2c3d4 \
  -H "Authorization: Bearer ak_your_api_key_here"
```
```

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

### Publish Site to Production

Source: https://www.tryflint.com/docs/guides/claude-ad-keyword-pages

Use this prompt to push the current site state to production after reviewing drafts.

```text
Publish the Reducto site to production.
```

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

### check_background_agent_status

Source: https://www.tryflint.com/docs/claude-mcp

Checks the status of a previously started background agent.

```APIDOC
## POST /check_background_agent_status

### Description
Checks the status of a previously started background agent.

### Method
POST

### Endpoint
`/check_background_agent_status`

### Parameters
#### Request Body
- **workflowId** (string) - Required - The ID returned from `run_background_agent`.

### Request Example
```json
{
  "workflowId": "workflow-abc-123"
}
```

### Response
#### Success Response (200)
- **status** (string) - The current status of the agent (e.g., 'running', 'completed', 'failed').
- **phase** (string) - The current phase of the agent's execution.
- **output** (string) - The output of the agent when completed.

#### Response Example
```json
{
  "status": "completed",
  "phase": "generating_content",
  "output": "Page content generated successfully."
}
```
```

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

### GET /api/v1/agent/tasks/{taskId}

Source: https://www.tryflint.com/docs/api

Retrieves the current status and output of an asynchronous task.

```APIDOC
## GET /api/v1/agent/tasks/{taskId}

### Description
Polls the status of an asynchronous task. Tasks can be in 'running', 'completed', or 'failed' states.

### Method
GET

### Endpoint
https://app.tryflint.com/api/v1/agent/tasks/{taskId}

### Parameters
#### Path Parameters
- **taskId** (string) - Required - The unique identifier of the task.

### Response
#### Success Response (200)
- **taskId** (string) - The task ID.
- **status** (string) - The current status ('running', 'completed', 'failed').
- **output** (object) - Present if completed; contains pagesCreated, pagesModified, and pagesDeleted arrays.
- **errorMessage** (string) - Present if failed; description of the error.

#### Response Example
{
  "taskId": "bg-550e8400-a1b2c3d4",
  "status": "completed",
  "output": {
    "pagesCreated": [
      {
        "slug": "/about",
        "previewUrl": "https://your-site-abc123.vercel.app/about",
        "editUrl": "https://app.tryflint.com/app/edit?siteId=...&page=about"
      }
    ],
    "pagesModified": [],
    "pagesDeleted": []
  }
}
```

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

### Set Nginx Startup Command

Source: https://www.tryflint.com/docs/azure-subpath-forwarding

Use this command in the App Service configuration to apply the custom Nginx settings during startup.

```bash
cp /home/site/nginx-custom.conf /etc/nginx/conf.d/flint-proxy.conf && service nginx reload && <your-normal-startup-command>
```

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

### Get Task Status Endpoint

Source: https://www.tryflint.com/docs/api

The endpoint used to retrieve the status of a specific task.

```text
GET /agent/tasks/{taskId}
```

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

### publish_site

Source: https://www.tryflint.com/docs/claude-mcp

Deploys a Flint site to production.

```APIDOC
## POST /publish_site

### Description
Deploys a Flint site to production. This resolves the latest changes on the staging branch and triggers a production deployment.

### Method
POST

### Endpoint
`/publish_site`

### Parameters
#### Request Body
- **siteId** (string (UUID)) - Required - The target site to publish.

### Request Example
```json
{
  "siteId": "site-uuid-123"
}
```

### Response
#### Success Response (200)
This endpoint typically returns a success status or confirmation message upon successful deployment.

#### Response Example
```json
{
  "message": "Site site-uuid-123 deployed to production successfully."
}
```
```

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

### Prompt to Create SEO-Optimized Template

Source: https://www.tryflint.com/docs/seo-landing-pages

Use this prompt to instruct Flint to generate a base template optimized for search engines. Ensure it includes key SEO elements like H1, H2s, FAQs, internal links, and CTAs.

```text
Create an SEO-optimized template with:
- H1 targeting primary keyword
- Structured H2 sections for key topics
- FAQ section at the bottom
- Internal linking opportunities
- Clear CTAs without disrupting content flow
```

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

### Configure IIS Rewrite Rules

Source: https://www.tryflint.com/docs/azure-subpath-forwarding

Add these rules to the web.config file within the system.webServer section to proxy requests to Flint.

```xml
<rewrite>
  <rules>
    <rule name="FlintProxy" stopProcessing="true">
      <match url="^pages/?(.*)"/>
      <conditions>
        <add input="{HTTP_HOST}" pattern="yoursite\.com"/>
      </conditions>
      <action type="Rewrite"
              url="https://yoursite.tryflint.com/{R:1}"/>
      <serverVariables>
        <set name="HTTP_X_FORWARDED_HOST" value="{HTTP_HOST}"/>
        <set name="HTTP_X_FORWARDED_PROTO" value="https"/>
      </serverVariables>
    </rule>
  </rules>
</rewrite>
```

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

### Configure MCP Server Scopes

Source: https://www.tryflint.com/docs/claude-code-mcp

Commands to add the Flint MCP server with different configuration scopes: local, user-wide, or project-shared.

```bash
# Local scope (default) - only you, only this project
claude mcp add flint --transport http https://mcp.tryflint.com/mcp

# User scope - available to you across all projects
claude mcp add flint --transport http --scope user https://mcp.tryflint.com/mcp

# Project scope - shared with everyone via .mcp.json
claude mcp add flint --transport http --scope project https://mcp.tryflint.com/mcp
```

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

### Vercel Configuration for Folder Routing

Source: https://www.tryflint.com/docs/setup

Configure Vercel to route traffic from a folder on your main domain to your Flint pages. This setup is useful for SEO and branding.

```json
{
  "rewrites": [
    {
      "source": "/lp",
      "destination": "https://yoursite.tryflint.com/lp"
    },
    {
      "source": "/lp/:match*",
      "destination": "https://yoursite.tryflint.com/lp/:match*"
    }
  ]
}
```

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

### Verify the Connection

Source: https://www.tryflint.com/docs/claude-code-mcp

Confirm that the Flint MCP server has been successfully added and configured in Claude Code.

```APIDOC
## Verify the Connection

### Description
Confirm that the Flint MCP server has been successfully added and configured in Claude Code.

### Commands
- **List all configured servers**:
  ```bash
  claude mcp list
  ```
  You should see `flint` in the list.

- **Get details for the Flint server**:
  ```bash
  claude mcp get flint
  ```
  This command provides more details about the configured Flint server.
```

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

### Publish Site Tool

Source: https://www.tryflint.com/docs/claude-mcp

Deploys a Flint site to production. Requires the site ID of the target site. This tool resolves the latest changes on the staging branch and triggers a production deployment, allowing for a quick transition from editing to a live site.

```json
{
  "tool_code": "publish_site",
  "parameters": {
    "siteId": "<site_uuid>"
  }
}
```

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

### Verify and Inspect MCP Connection

Source: https://www.tryflint.com/docs/claude-code-mcp

Commands to list configured MCP servers and retrieve specific details for the Flint server.

```bash
claude mcp list
```

```bash
claude mcp get flint
```

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

### Generate Keyword Metrics for Comparison Pages

Source: https://www.tryflint.com/docs/guides/claude-ad-keyword-pages

Use this prompt to generate a table of 'vs' keywords with estimated search volume, CPC, and competition scores.

```text
Based on the competitor list, come up with each
"vs" keyword (e.g. "[Your Product] vs [Competitor]")
with estimated search volume, CPC, and competition
score. Present the top 5 in a table view.
```

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

### Connect to Azure VM via SSH

Source: https://www.tryflint.com/docs/azure-subpath-forwarding

Establish a secure shell connection to the virtual machine using the provided username and IP address.

```bash
ssh your-username@YOUR_VM_IP
```

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

### Create Task Endpoint

Source: https://www.tryflint.com/docs/api

The endpoint used to initiate a background agent task.

```text
POST /agent/tasks
```

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

### Troubleshooting

Source: https://www.tryflint.com/docs/claude-code-mcp

Common issues and their solutions for the Claude Code MCP Integration with Flint.

```APIDOC
## Troubleshooting

### Connection not working?
1. Verify the server URL is entered exactly as shown in the setup instructions.
2. Try removing and re-adding the server: run `claude mcp remove flint` and then add it again using the `claude mcp add` command.

### Authentication issues?
1. Run `/mcp` in Claude Code and select "Authenticate" for Flint.
2. Complete the browser login flow.
3. If the browser redirect fails, copy the callback URL from your browser and paste it into Claude Code.

### Pages not appearing?
1. Check that the task completed successfully in Claude Code's response.
2. Refresh your Flint dashboard to see newly created pages.
3. Verify you have the correct Flint site selected for the operation.
```

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

### Generate A/B Test Variants with Statsig

Source: https://www.tryflint.com/docs/abtesting

Use this prompt to instruct Flint to create A/B test variants for a specific page, including changing header text and applying CSS classes. Ensure Statsig is loaded via a script tag and accessible via `window.statsig`.

```text
Create a variant of the / page called /variant-hero with Statsig A/B testing for the hero section.

Requirements:
1. The /variant-hero should change header text from "Your company's always-on website agent" to "Landing pages that build themselves"

2. Implement Statsig A/B testing for the hero copy using this logic: [Add any specific implementation JS script from your experimentation platform]

3. Add these classes to the hero elements:
Headline: class="hero-headline"
Subheadline: class="hero-subheadline"
CTA Button: class="cta-button"
Statsig is already loaded via script tag, so just use window.statsig directly.
```

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

### Edit Nginx Configuration on VM

Source: https://www.tryflint.com/docs/azure-subpath-forwarding

Open the Nginx site configuration file for editing to add the proxy rules.

```bash
sudo nano /etc/nginx/sites-available/yoursite.com
```

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

### Create Custom Nginx Configuration

Source: https://www.tryflint.com/docs/azure-subpath-forwarding

Define the proxy location block for Nginx on Linux App Services to route traffic to the Flint subdomain.

```bash
cat > /home/site/nginx-custom.conf << 'NGINX_CONF'
location /pages/ {
    proxy_pass https://yoursite.tryflint.com/;
    proxy_set_header Host yoursite.tryflint.com;
    proxy_set_header X-Forwarded-Host $http_host;
    proxy_set_header X-Forwarded-Proto https;
    proxy_set_header X-Real-IP $remote_addr;
    proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
    proxy_ssl_server_name on;
}

location = /pages {
    return 301 /pages/;
}
NGINX_CONF
```

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

### Authenticate with Flint

Source: https://www.tryflint.com/docs/claude-code-mcp

After adding the server, authenticate with your Flint account through Claude Code. This process involves opening the MCP menu, selecting Flint, and completing a browser-based sign-in.

```APIDOC
## Authenticate with Flint

### Description
After adding the Flint MCP server, you need to authenticate your Flint account within Claude Code. This allows Claude Code to create and modify pages on your behalf.

### Steps
1. **Open the MCP menu in Claude Code**
   Use the command `/mcp` in your Claude Code terminal.

2. **Select Flint and authenticate**
   From the menu, choose Flint and follow the prompts to sign in to your Flint account via your browser. This authorizes Claude Code.

### Troubleshooting Authentication Issues
- Run `/mcp` in Claude Code and select "Authenticate" for Flint.
- Complete the browser login flow.
- If the browser redirect fails, copy the callback URL from your browser and paste it into Claude Code.
```

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

### list_sites

Source: https://www.tryflint.com/docs/claude-mcp

Lists all Flint sites accessible to the authenticated account.

```APIDOC
## GET /list_sites

### Description
Lists all Flint sites accessible to your account. This is useful for discovering your site IDs.

### Method
GET

### Endpoint
`/list_sites`

### Parameters
No parameters required.

### Response
#### Success Response (200)
- **sites** (array) - A list of sites.
  - **name** (string) - The name of the site.
  - **siteId** (string (UUID)) - The unique identifier for the site.
  - **organizationId** (string) - The identifier for the organization.
  - **organizationName** (string) - The name of the organization.

#### Response Example
```json
{
  "sites": [
    {
      "name": "My Awesome Site",
      "siteId": "site-uuid-123",
      "organizationId": "org-uuid-456",
      "organizationName": "My Company"
    }
  ]
}
```
```

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

### List Sites Tool

Source: https://www.tryflint.com/docs/claude-mcp

Lists all Flint sites accessible to your account. No parameters are required. This tool is useful for discovering site IDs directly within a Claude conversation.

```json
{
  "tool_code": "list_sites",
  "parameters": {}
}
```

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

### Test Nginx configuration

Source: https://www.tryflint.com/docs/azure-subpath-forwarding

After modifying your Nginx configuration, test its syntax and validity before reloading the service.

```bash
sudo nginx -t
```

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

### Generate Event Registration Page Prompt

Source: https://www.tryflint.com/docs/guides/event-landing-pages

Use this prompt to instruct Flint to create an event registration page. Customize the bracketed information with your specific event details.

```text
Create an event registration page for [Event Name] on [Date].
Include: hero with countdown timer, 3 key benefits, featured speakers section, agenda overview, testimonials, and a sticky registration CTA.

```

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

### Managing the Flint MCP Server

Source: https://www.tryflint.com/docs/claude-code-mcp

Commands to manage the Flint MCP server configuration within Claude Code.

```APIDOC
## Managing the Server

Use these commands to manage your Flint MCP server configuration in Claude Code:

### List all configured servers
```bash
claude mcp list
```

### Get details for the Flint server
```bash
claude mcp get flint
```

### Remove the Flint server
```bash
claude mcp remove flint
```

### Check server status in Claude Code
Use the command `/mcp` in Claude Code.
```

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

### Manage Flint MCP Server

Source: https://www.tryflint.com/docs/claude-code-mcp

Common administrative commands for listing, inspecting, removing, and checking the status of the Flint MCP server.

```bash
# List all configured servers
claude mcp list

# Get details for the Flint server
claude mcp get flint

# Remove the Flint server
claude mcp remove flint

# Check server status in Claude Code
/mcp
```

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

### Flint MCP Tools

Source: https://www.tryflint.com/docs/claude-code-mcp

The Flint MCP server exposes several tools that Claude Code can utilize to manage your Flint sites.

```APIDOC
## Available Tools

The Flint MCP server exposes four tools that Claude Code can use to interact with your Flint sites:

### `run_background_agent`

#### Description
Start a background AI agent on a Flint site. The agent executes a prompt to modify your website.

#### Parameters
- **siteId** (string (UUID)) - Required - The UUID of the Flint site to modify.
- **prompt** (string) - Required - What the background agent should do.

#### Returns
A workflow ID that can be used to check the status of the agent.

### `check_background_agent_status`

#### Description
Check the status of a previously started background agent.

#### Parameters
- **workflowId** (string) - Required - The ID returned from `run_background_agent`.

#### Returns
The current status, phase, and output when the agent has completed.

### `list_sites`

#### Description
Lists all Flint sites accessible to your account.

#### Parameters
No parameters required.

#### Returns
A list of sites with their name, site ID, organization ID, and organization name. This is useful for discovering your site IDs without leaving the conversation.

### `publish_site`

#### Description
Deploys a Flint site to production.

#### Parameters
- **siteId** (string (UUID)) - Required - The target site to publish.

#### Returns
Resolves the latest changes on the staging branch and triggers a production deployment.
```

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

### Reload Nginx service

Source: https://www.tryflint.com/docs/azure-subpath-forwarding

Apply the Nginx configuration changes by reloading the Nginx service.

```bash
sudo systemctl reload nginx
```

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

### Flint MCP Server URL

Source: https://www.tryflint.com/docs/claude-mcp

Paste this URL when adding a custom connector in Claude.

```text
https://mcp.tryflint.com/mcp
```

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

### Add Flint to Claude Code

Source: https://www.tryflint.com/docs/claude-code-mcp

Run this command in your terminal to add Flint as an MCP server. This command adds Flint as a remote HTTP MCP server that Claude Code can use to build and manage your landing pages.

```APIDOC
## Add Flint to Claude Code

### Description
Run this command in your terminal to add Flint as an MCP server. This command adds Flint as a remote HTTP MCP server that Claude Code can use to build and manage your landing pages.

### Method
Bash

### Endpoint
N/A

### Parameters
#### Query Parameters
- **transport** (string) - Required - Specifies the transport protocol, typically 'http'.
- **url** (string) - Required - The URL of the Flint MCP server, e.g., `https://mcp.tryflint.com/mcp`.
- **scope** (string) - Optional - Controls where the MCP server configuration is stored. Options: `local` (default), `user`, `project`.

### Request Example
```bash
claude mcp add flint --transport http https://mcp.tryflint.com/mcp
```

### Request Example with Scope
```bash
# User scope - available to you across all projects
claude mcp add flint --transport http --scope user https://mcp.tryflint.com/mcp

# Project scope - shared with everyone via .mcp.json
claude mcp add flint --transport http --scope project https://mcp.tryflint.com/mcp
```
```

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

### In Progress Task Status

Source: https://www.tryflint.com/docs/api

This JSON indicates that a task is currently running.

```json
{
  "taskId": "bg-550e8400-a1b2c3d4",
  "status": "running"
}
```

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

### Agent Tasks API

Source: https://www.tryflint.com/docs

Interact with Agent Tasks through the Flint API.

```APIDOC
## Agent Tasks API

### Description
This API allows you to programmatically manage and interact with Agent Tasks within Flint.

### Method
GET, POST, PUT, DELETE (Specific methods depend on the task)

### Endpoint
`/websites/tryflint/api/agent-tasks`

### Parameters
(Specific parameters will be detailed in sub-sections for each task)

### Request Example
(Example request bodies will vary based on the specific agent task)

### Response
#### Success Response (200)
- **status** (string) - The status of the agent task.
- **result** (object) - The result of the agent task execution.

#### Response Example
```json
{
  "status": "completed",
  "result": {
    "message": "Task executed successfully."
  }
}
```
```

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

### Completed Task Status and Output

Source: https://www.tryflint.com/docs/api

This JSON represents a successfully completed task, including details about created, modified, and deleted pages.

```json
{
  "taskId": "bg-550e8400-a1b2c3d4",
  "status": "completed",
  "output": {
    "pagesCreated": [
      {
        "slug": "/about",
        "previewUrl": "https://your-site-abc123.vercel.app/about",
        "editUrl": "https://app.tryflint.com/app/edit?siteId=...&page=about"
      }
    ],
    "pagesModified": [],
    "pagesDeleted": []
  }
}
```

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

### Cloudflare Worker for Webflow Routing

Source: https://www.tryflint.com/docs/webflow-subpath-forwarding

This script routes requests for the /lp path to Flint and all other requests to your main Webflow site. Ensure you replace placeholders like [SUBDOMAIN], [YOUR_DOMAIN], and [LANDING_DOMAIN] with your actual values.

```javascript
addEventListener("fetch", (event) => {
  event.respondWith(handleRequest(event.request));
});

async function handleRequest(request) {
  try {
    const urlObject = new URL(request.url);

    // If the request is to the /lp folder
    if (/^\/lp/.test(urlObject.pathname)) {
      // Proxy to Flint
      const FLINT_URL = "[SUBDOMAIN].tryflint.com";
      const CUSTOM_URL = "[YOUR_DOMAIN]";
      let url = new URL(request.url);
      url.hostname = FLINT_URL;
      let proxyRequest = new Request(url, request);
      proxyRequest.headers.set("Host", FLINT_URL);
      proxyRequest.headers.set("X-Forwarded-Host", CUSTOM_URL);
      proxyRequest.headers.set("X-Forwarded-Proto", "https");
      return await fetch(proxyRequest);
    }

    // Route everything else to main site
    const MAIN_SITE_URL = "[LANDING_DOMAIN]";
    if (MAIN_SITE_URL && MAIN_SITE_URL !== "[LANDING_DOMAIN]") {
      let mainSiteUrl = new URL(request.url);
      mainSiteUrl.hostname = MAIN_SITE_URL;
      return await fetch(mainSiteUrl, {
        method: request.method,
        headers: request.headers,
        body: request.body,
      });
    }
  } catch (error) {
    // If no action found, serve the regular request
    return await fetch(request);
  }
}
```

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

### Configure URL Rewrite Rule for Flint

Source: https://www.tryflint.com/docs/azure-subpath-forwarding

This rule set strips the `/pages` prefix from the request URL before forwarding it to Flint. Ensure this rule set is associated with your route for it to take effect.

```azure-cli
az network front-door rule-set create --name flint-rewrite --resource-group <your-resource-group>
az network front-door rule-set rule add --rule-set-name flint-rewrite --resource-group <your-resource-group> --name strip-subpath --action-type "UrlRewrite" --source-pattern "/pages/" --destination "/" --preserve-unmatched-path true --condition-type "RequestUrl" --operator "BeginsWith" --match-value "/pages/"
```

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

### Prompt to Add Internal Linking Section

Source: https://www.tryflint.com/docs/seo-landing-pages

Use this prompt to add a structured internal linking section to an existing page. It specifies the number and type of links to include, emphasizing descriptive anchor text.

```text
Add an internal linking section to this page:
- 3 related blog posts
- 2 related solution pages
- Links to parent category page
Use descriptive anchor text with relevant keywords.
```

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

### Claude MCP Integration API

Source: https://www.tryflint.com/docs

Integrate with Claude MCP using the Flint API.

```APIDOC
## Claude MCP Integration API

### Description
This API facilitates integration with Claude MCP, enabling advanced functionalities.

### Method
GET, POST (Specific methods depend on the integration point)

### Endpoint
`/websites/tryflint/api/claude-mcp`

### Parameters
(Specific parameters will be detailed based on the integration use case)

### Request Example
(Example request bodies will vary based on the Claude MCP integration)

### Response
#### Success Response (200)
- **data** (object) - The response data from the Claude MCP integration.

#### Response Example
```json
{
  "data": {
    "processed_text": "This is the processed text from Claude MCP."
  }
}
```
```
Tokens:
Raw