Client Foundation
Before you create an API key or install an MCP server, decide what kind of Site Clinic implementation you are actually starting. A finished website, an API integration, a monitoring account, and an agent-assisted workflow have different requirements and proof artifacts.
Choose the right path
Use this path when the goal is a deployed website with content, domain routing, monitoring, and launch proof. The API can support checks, but it does not replace the website project itself.
Use this path when your app, agency tooling, or internal system needs ADA, health, AI visibility, usage, or report data through authenticated API requests.
Use this path when a site already exists and the job is recurring scans, alerts, dashboard review, fixes, and verified improvement evidence.
Use this path only when your team already works through an MCP-capable coding assistant and wants curated Site Clinic tools exposed through that client.
Minimum requirements
A Site Clinic account starts the current public monitoring and developer/API layers through Stripe Checkout. Public plans begin with a 30-day trial before paid billing continues.
A finished website still needs business name, offer, audience, pages, copy direction, brand assets, contact details, legal links, and launch domain decisions.
Developer/API work assumes someone can run a project locally or in a hosted build environment, store secrets, deploy code, and read request logs.
Each implementation should leave evidence: deployed URL, dashboard access, scan results, request IDs, usage records, and before/after verification where applicable.
Website truth
Site Clinic can support finished website delivery, but the public API is not a magic website generator. The API returns structured analysis and workflow data. A real website still requires a project, content, design choices, deployment, domain setup, and monitoring configuration.
Web Builder assisted delivery is the correct path when the client outcome is a complete website. Developer docs are the correct path when the client outcome is an integration that calls Site Clinic services from their own code.
The 30-day trial is for recurring Site Clinic access such as monitoring and developer/API usage. A custom finished website build is a separate delivery scope; otherwise the build work would become a free one-time deliverable that survives cancellation while the recurring product turns off.
AI assistance on the website itself is a future upgrade. The current foundation should explain the available product surfaces clearly before asking a client to choose tools they may not need.
30-day trial and delivery boundary
The public developer account flow is Stripe-backed and includes a 30-day trial for the selected developer plan. During the trial, the account can create test and live API keys and use the included plan quotas.
If a client wants Site Clinic to build a complete website during trial, the boundary must be explicit before work begins. Customer-owned domain, repo, and public pages can remain with the customer. The paid subscription is the revocable operating system around the website: monitoring, evidence, API/MCP access, scheduler-owned workflows, Blog Writer, reports, alerts, and ongoing proof.
API keys are for authenticated programmatic access. They should be stored as secrets, separated by environment, and rotated when ownership or exposure risk changes.
MCP access uses the same product truth as the API: it is an agent surface over promoted Site Clinic capabilities, not a separate promise that every internal tool is public or production-ready.
Completion standard
A live URL, domain routing, working contact path, core pages, metadata, accessibility pass, monitoring setup, and launch notes.
A successful authenticated request, error handling, quota awareness, stored request IDs, usage dashboard visibility, and a production-secret plan.
A tracked site, recurring scan cadence, dashboard proof, alert path, and a documented fix-and-verify loop.
A configured MCP client, documented tool list, API-key authentication, and proof that the invoked tool produced the expected Site Clinic evidence.