Build with AI

View as Markdown

Sudomimus publishes a Markdown version of each documentation page for coding assistants. Give your assistant the pages for your integration path. It can fetch more pages when the implementation needs them.

Choose the relevant pages

If you have not chosen an integration path, start with Choose an integration path. Then give your assistant the matching flow page:

Integration path	Read first	Exact API contract
Connect CALLBACK for a web application	Connect flow and Return rules	Connect API and Session API
Connect STATUS_POLL with a trusted signing companion	Connect browser polling and Return rules	Connect API and Session API
Device authorization for a public client	Device authorization flow and Polling and errors	Device API and Session API
OIDC for a relying party	OIDC flow	OIDC discovery at `https://oidc.sudomimus.com/.well-known/openid-configuration`
Native direct-issue with a platform credential	Native flows	Native API and Session API

Add .md to a documentation page URL to give an assistant just that page. For example, use https://docs.sudomimus.com/en-us/connect/flow.md for Connect. Use the matching API reference when exact request, response, or error shapes matter. The prompt builder selects the pages for its supported flows.

A starting prompt

Paste this into Claude, Cursor, ChatGPT, or any AI tool with web access, filling in the angle-bracketed parts:

Read https://docs.sudomimus.com/en-us/ai/overview.md. Choose the matching
integration path and read only its flow and API contracts. Integrate Sudomimus
authentication into my application.

Stack: <your stack, e.g. Next.js 14 + Drizzle + Postgres>
Auth methods: <e.g. passkeys + email OTP>
Application anchor: <your-application>
Callback URL: <your callback URL>

You still need to supply the surrounding project context, configure the application, review the generated code, and verify the flow against the published contract. Use llms-full.txt only when the task needs broad documentation context across integration paths.

If the assistant needs to operate your Sudomimus account or developer settings while it works, give it the Sudomimus CLI instead of browser instructions. The CLI exposes login, session, and account commands in a form an agent can call directly.

Where to go next

What to feed your AI — the machine-readable endpoints, per-page Markdown, and the OpenAPI reference.
Driving your AI assistant — tool-specific tips for Claude, Cursor, ChatGPT and others, and how to get better integration code out of them.
Sudomimus CLI for AI agents — the command-line tool for account checks, local automation, and future developer operations.
Sudomimus MCP server — connect a local MCP host to With through the CLI session.

Why this works

The documentation covers the choice between Connect, OIDC, and native direct-issue; their protocol flows; the shared identity model; and the three-layer rule model. Giving an assistant the relevant pages lets it work from the same published material as a human integrator. Re-fetch pages when the contract may have changed, and treat generated code as an implementation draft that still needs project-specific review and testing.