·7 min read
ByGabriel CA·Kraftwire Software

How to Use the SimplyScan MCP Server for Cursor and Windsurf

Illustration for How to Use the SimplyScan MCP Server for Cursor and Windsurf

Connect a compatible coding assistant to SimplyScan to invoke supported checks from your development workflow. Review the returned findings and coverage, then test proposed changes. A URL scan needs a reachable deployment; use a test deployment when you need observations before a production release.

Why Use the SimplyScan MCP Server for AI Development?

An editor integration can reduce switching between code and scan results. Choose tools according to their actual coverage and the inputs they inspect. Source review, dependency checks, and deployed-site observations address different questions.

These live figures describe all scanned sites. They do not establish prevalence for an AI-built subgroup, compare development methods, or attribute findings to AI generation.

32% of the 127 live sites SimplyScan has scanned (41 sites) have at least one high or critical security issue.

Exposed credentials and unintended public responses are risks to investigate; the aggregate figure does not show their individual prevalence. The scan_website tool inspects a deployed URL. It does not verify every Supabase control, inspect all local code, or test signed-in cross-user access.

How Do You Install the SimplyScan MCP Server in Cursor?

Use the Cursor MCP documentation for the installed client version. Remote-server setup differs from starting a local command.

Step 1

Open Cursor’s MCP settings or configuration using the instructions for your installed version.

Step 2

Choose the project or user configuration scope you intend to use.

Step 3

Add a remote MCP server entry.

Step 4

Name the server SimplyScan.

Step 5

Set the type to a remote (URL) server.

Step 6

Enter the server URL: https://api.simplyscan.io/mcp.

Once connected, inspect the exposed tools and invoke a relevant check. Secret-pattern checks and AEO signals have different inputs and limits; neither proves that a suggested fix is complete.

How Do You Configure SimplyScan for Windsurf?

Use the configuration for your installed agent. The Windsurf/Cascade documentation describes remote MCP entries; other agents or newer clients may use a different configuration file.

Step 1

Open the Windsurf MCP configuration file (mcp_config.json, reachable from the MCP settings UI).

Step 2

Add the SimplyScan remote entry to the mcpServers object, merging it with any existing servers:

{
  "mcpServers": {
    "simplyscan": {
      "serverUrl": "https://api.simplyscan.io/mcp"
    }
  }
}

Step 3

Save and refresh the MCP connection as required by your client. Confirm that the expected tools appear before invoking them.

Ask the agent to invoke the relevant security checks and inspect the returned coverage. A new route or deployment does not automatically trigger a header check unless that workflow is configured.

What Can the SimplyScan MCP Server Detect?

The server exposes SimplyScan tools through MCP. A free URL scan typically takes about 30 seconds; tool inputs, limits, and timing vary.

10% of the 127 live sites SimplyScan has scanned (13 sites) have at least one medium-severity architecture issue, such as a library with a known vulnerability or no rate limiting.

The tool results can include the supported signals below. The aggregate architecture figure does not establish that a particular integration has these findings.

Security and Auth Signals

Supported checks inspect recognizable exposed credentials, selected client-code patterns, headers, and anonymous backend responses. A URL scan does not verify all RLS policies or authenticated flows. Use the Lovable guide alongside direct permission tests.

Performance and Speed

80% of the 127 live sites SimplyScan has scanned (101 sites) have at least one medium-severity speed issue, such as render-blocking scripts or no route-based code splitting.

URL results can include supported loading and asset findings, with CrUX field metrics where available. Review the measured evidence and missing coverage, then test any proposed refactor under comparable conditions.

Compliance and Domain Signals

Supported tools inspect privacy-related HTML patterns, selected SPF/DKIM/DMARC DNS records, and TLS/certificate observations. These do not establish legal compliance or successful email authentication. Use the SSL checker to inspect supported certificate and transport results.

How Do You Check Client Compatibility?

Confirm the transport and authentication settings supported by your installed client and the service. Test connection initialization and tool discovery, then inspect any errors.

The supported endpoint is listed on the SimplyScan MCP page. Do not infer support for every new protocol feature from a successful basic connection.

Can the AI Automatically Fix Security Issues?

A useful way to use the SimplyScan MCP server is to create a feedback loop.

You can prompt your IDE with: "Scan my staging URL and fix any high-severity security issues you find."

Use the following review sequence with the agent:

Step 1

Invoke the URL check on the intended deployment.

Step 2

Inspect the returned report and coverage, including suspected exposed-key findings.

Step 3

Trace a confirmed exposed value to the relevant code or deployment configuration and investigate misuse. Revoke the exposed credential.

Step 4

Deploy a replacement through protected server configuration. Exclude sensitive .env files from version control with .gitignore, and check that the framework does not publish the value to the client.

Step 5

Rescan the changed deployment and test the affected behavior directly. A missing finding is meaningful only within the checks and resources covered.

Use MCP results alongside static analysis and manual review. An editor feedback loop does not replace those methods or prove that an automated code change is correct.

Is the SimplyScan MCP Server Free to Use?

The MCP server can run a free URL scan without signup, subject to usage limits. It grades 8 dimensions using supported checks. Pro tools and full reports have separate entitlement requirements; the local-code tool requires a Pro API key and uploads selected files for analysis.

Pro Monitoring includes scan credits, scheduled rescans, uptime monitoring, status pages, and supported Slack/GitHub/Linear integrations. Evaluate the features against your workflow; they do not provide continuous verification of every application control.

How Does This Compare to Other MCP Servers?

Compare MCP servers by the tools, permissions, inputs, and outputs they expose. Repository access, browser automation, database operations, and bounded scan observations are different capabilities.

Using a database tool alongside SimplyScan MCP does not establish a secure configuration. Review proposed policies, deploy them deliberately, and test allowed and denied requests with appropriate accounts. Anonymous URL observations do not prove that RLS blocks every unauthorized user.

Summary of Best Practices for AI Security

Choose relevant checks for changes to integrations, routes, and deployment settings:

Review findings before accepting changes from the agent. Record untested controls and perform the relevant manual or authenticated checks.

Start with a free URL scan at SimplyScan to inspect supported public-facing signals. Read the report’s coverage alongside its findings.

Pro Monitoring costs $24/month.

Sources

Every figure above was checked against the page it cites · here they are, so you can check them too.

Show 2 sources
  1. ayautomate.com › best-mcp-servers-2026

    GitHub for repos and PRs, Filesystem for local files, Playwright for browser control, Supabase and Postgres MCP Pro for databases, Cloudflare for edge infra, Stripe for payments, Sentry for errors, Notion for docs, and Context7 for live library docs.

  2. blog.modelcontextprotocol.io › 2026-07-28

    2026-07-28

Frequently asked questions

Use Cursor’s MCP configuration for your installed version and add a remote server at https://api.simplyscan.io/mcp. Confirm that tool discovery succeeds, then invoke the relevant check and inspect its returned findings and coverage.
Supported tools report recognizable secret patterns, selected headers and public-access responses, loading signals, and other bounded observations. A URL scan does not verify every RLS policy, sign in as different users, or establish that an application is secure.
The MCP server supports free URL checks without signup, subject to usage limits. Some tools and full reports require appropriate entitlement. The Pro local-code tool uploads selected source files for analysis and requires a Pro API key.
Use the configuration supported by your installed Windsurf/Cascade client. Add a remote entry, refresh the connection, and verify tool discovery. The agent can invoke checks, but automatic triggers and edits depend on the workflow and client permissions.
Check the transport, authentication, and protocol compatibility documented for your installed client and the service. Test initialization and tool discovery. This guide does not claim support for every feature of a dated protocol revision.
SimplyScan provides supported deployed-site observations and selected source checks through its tools. Compare these with the actual source, dependency, and other capabilities of each alternative. A URL scan does not verify all database policies, and the tools can serve complementary roles.

Is your app secure? Run a free security scan in 30 seconds · free account required.

Scan your app free →