CheckMyLaunch MCP server
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@CheckMyLaunch MCP server scan my GitHub repo for launch issues"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
CheckMyLaunch MCP server
Lets an AI coding tool (Claude Code, Claude Desktop, Cursor, etc.) run a CheckMyLaunch scan and read back findings + repair prompts without leaving the chat.
This is a paid feature. Every tool call is gated behind a CheckMyLaunch API key from a paid account (one-time Launch Check purchase or the Continuous Guard subscription) — get one at checkmylaunch.com/account after signing in and upgrading. Without a valid key configured, every tool refuses with a message pointing you there instead of running.
Tools
scan_repo
{ repoUrl }— scans a public GitHub repo.scan_url
{ url }— scans a live site.get_repair_prompt
{ scanId, findingId }— fetches the paste-back fix prompt for one finding.
Related MCP server: aria51 MCP Server
Setup
Sign in at checkmylaunch.com, upgrade to a paid plan, then go to your account page and generate an API key.
Add this to your MCP client's config:
{
"mcpServers": {
"checkmylaunch": {
"command": "npx",
"args": ["github:matdtools-code/checkmylaunch-mcp"],
"env": { "CHECKMYLAUNCH_API_KEY": "cml_live_..." }
}
}
}Claude Code: claude mcp add or edit .claude/mcp.json /
~/.claude.json. Claude Desktop / Cursor: same JSON shape in their
respective MCP config files.
Try it locally
npm install
CHECKMYLAUNCH_API_KEY=cml_live_... node test/smoke.jsConfig
CHECKMYLAUNCH_API_KEY— required. From your CheckMyLaunch account page.CHECKMYLAUNCH_API_URL— override the API base URL (defaults tohttps://www.checkmylaunch.com); useful for pointing atlocalhost:8899during development.
How the gate works
Every tool call first verifies the key against GET /api/auth/api-key on
the CheckMyLaunch server, which only resolves for a signed-in, paid account
(see server/src/routes/auth.js in the main app repo). A missing, invalid,
or free-plan key stops the call before any scan runs.
This doesn't (and can't) prevent someone from calling CheckMyLaunch's public
/api/scan/public-github / /api/scan/url endpoints directly without this
plugin — those intentionally stay open, capped at a small free weekly quota
with truncated findings, as the website's own no-signup teaser. What the key
unlocks is the difference that actually matters for real use: uncapped
findings and no weekly quota, from inside your coding tool.
Available Tools
3 toolsget_repair_promptA
Get a ready-to-use repair prompt for one finding from a previous scan_repo or scan_url result. The returned prompt is safe to run as-is in a coding tool: it names the exact fix, forbids unrelated changes, and never deploys or rotates secrets automatically.
Inputs: scanId (required), findingId (required).
scan_repoA
Run a CheckMyLaunch readiness scan on a public GitHub repository. Finds exposed secrets, security risks, deployment problems and launch blockers, and returns a readiness score plus a findings list. Use this after making changes to a project to check it is safe to ship. Only works on public repos reachable at github.com// — private repos need the signed-in flow at checkmylaunch.com.
Inputs: repoUrl (required).
scan_urlA
Run a CheckMyLaunch readiness scan on a live website URL. Checks security headers, cookie flags, TLS, exposed sensitive paths, mixed content, broken links and metadata, and returns a readiness score plus a findings list. Use this to check a deployed site before announcing a launch.
Inputs: url (required).
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 3 tools
Each tool has a distinct purpose: scan_repo scans a repository, scan_url scans a live URL, and get_repair_prompt provides repair instructions. No overlap.
All tools follow a consistent verb_noun snake_case pattern: scan_repo, scan_url, get_repair_prompt. Highly predictable.
Three tools is a reasonable minimal set for a security scanning server. While it could be expanded, the count is appropriate for the core functionality.
The set covers scanning code (repo) and live sites (URL), plus remediation (repair prompt). A minor gap is lack of result listing or management, but core workflows are covered.
Related MCP Connectors
Scan websites and code for security, performance and SEO issues from your AI assistant.
Scan a website for AI search and agent readiness, then get a fix checklist for your agent.
Run WCAG accessibility scans on websites, list each issue with its element, and triage from chat.
Scan URLs for WCAG 2.1 violations, generate AI fixes, and produce VPAT 2.5 compliance reports.
Related MCP Servers
AlicenseBqualityBmaintenanceEnables deep security auditing of web applications directly from AI IDEs including Cursor and Claude Code. Scans URLs for vulnerabilities, returns security scores with SHIP/BLOCK verdicts, and provides specific fix prompts for remediation.3243 npmMIT- AlicenseNot gradedqualityBmaintenanceEnables AI coding assistants to test web accessibility by scanning URLs, detecting violations, and running focused audits on keyboard navigation, screen reader compatibility, and WCAG criteria — all within the assistant's loop.MIT
- AlicenseAqualityBmaintenanceEnables AI coding assistants to validate HTML/CSS markup using W3C APIs, perform technical SEO audits, check broken links, and validate JSON-LD schemas directly in local workspaces.8802 npm10MIT
- AlicenseNot gradedqualityBmaintenanceEnables AI coding assistants to audit a site's technical SEO and AI-search readiness — including localhost dev servers — and retrieve rule explanations, fix templates, and generated llms.txt files so the assistant can edit the repo to resolve what it finds.6 npmMIT