Launch Check
Free launch check of this project: is it ready to go live and safe? Rule-based checks on secret keys in page code and script files, data that anyone can read, mixed content, search engine indexing, pages hidden from search engines, own domain, SEO titles and descriptions, forms that store or send nothing, outside backends, template placeholders and page quality (headings, mobile viewport, alt text). Call run, then show the owner each open item: status action needs work, review asks the owner to confirm, upgrade needs a paid plan. Every open item has detail for the owner and ai_instruction with the steps to fix it. No AI, no changes to the site.
Try it with your AI assistant
Once WebsitePublisher is connected to your AI assistant, paste one of these into the chat.
@WebsitePublisher is my site ready to go live? Run a launch check@WebsitePublisher check my site for security problems before launchWith WebsitePublisher, is my site ready to go live? Run a launch checkWith WebsitePublisher, check my site for security problems before launchWith WebsitePublisher, is my site ready to go live? Run a launch checkWith WebsitePublisher, check my site for security problems before launchEndpoints (3)
Run the launch check now and store the run. Returns check_id, score (0-100), summary (counts per status, fixable = actions a fix can solve), pages_inspected, pages_truncated and items. Each item: id, category (security, launch, function, quality), status (ok, action, review, upgrade), title; open items also severity, fix_kind (auto, ai, human, upgrade), detail, ai_instruction and the pages or forms involved. Within a few minutes of the last run it returns that run (reused true, next_run_after).
No input parameters required.
One stored launch check of this project: the given check_id, or the latest. Same fields as run. Read-only.
| Field | Type | Required | Description |
|---|---|---|---|
check_id |
integer | No | A check_id from run or list. Default: the latest run. |
Earlier launch checks of this project, newest first, without their items: check_id, created_at, trigger, plan_tier, score, action_count, fixable_count, pages_inspected. Read-only.
| Field | Type | Required | Description |
|---|---|---|---|
limit |
integer | No | Max runs, default 10, max 50. |
MCP Tool Names
When using this integration through an AI assistant (Claude, ChatGPT, Cursor, etc.), the endpoints are available as MCP tools:
| Endpoint | MCP Tool Name |
|---|---|
| run | launch_check_run |
| get | launch_check_get |
| list | launch_check_list |
Website