Skip to content

AI-Assisted Integration

anybill actively supports developers who integrate the Frontend SDK with the help of AI coding agents such as Cursor, Claude Code, GitHub Copilot and Codex. The primary audience is developer teams at retail merchants and integration partners who want to move from first contact to a working integration with fewer round-trips, less manual reading, and less trial-and-error.

Why we provide this

We treat Developer Experience as part of the product. If developers rely on AI coding agents day-to-day, the documentation has to be accessible to those agents as well, not only to humans browsing HTML. We therefore ship the documentation in a machine-readable shape, make the integration patterns explicit, and expose them at stable URLs that any AI tool can reference. It is part of how we build anybill, not a side project.

What we provide

Machine-readable documentation (llms.txt)

llms.txt is an open standard for publishing a curated index of documentation that is optimized for large language models. Our index lives at:

It lists every relevant SDK and API page with a one-line description and a direct link to the Markdown version.

Usage example for Cursor: open Settings > Features > Docs > Add new doc and paste the URL above. Cursor will index the referenced pages and make them available in chat and inline suggestions.

Clean Markdown versions of every page

Every documentation page is also served as a plain Markdown file at the same path with an .md suffix. For example:

These files contain the same content as the rendered HTML pages but without HTML boilerplate, which makes them ideal as context for AI coding agents.

AGENTS.md starter file

AGENTS.md is an open convention for project-level instructions aimed at AI coding agents. It is supported by Cursor, OpenAI Codex, Zed and other tools. We provide a ready-to-use starter file that captures the anybill authentication model, the SDK initialization contract, common pitfalls and out-of-scope actions.

Download it and place it at the root of your project:

The full content of the file:

How to use it with your AI tool

Cursor

  1. Open Settings > Features > Docs > Add new doc and add https://developer.anybill.de/llms.txt.
  2. Download anybill-sdk-AGENTS.md and save it as AGENTS.md at the root of your project.
  3. Prompt Cursor, for example:
Integrate the anybill Android SDK into this app using the linked user
account flow. Follow the anybill docs and the rules in AGENTS.md.

Claude Code

  1. Save anybill-sdk-AGENTS.md as CLAUDE.md at the root of your project, or create a symlink from CLAUDE.md to AGENTS.md.
  2. Reference the documentation via /add-dir for local copies or pass the URL https://developer.anybill.de/llms.txt directly in your prompt.
  3. Prompt Claude Code, for example:
Read https://developer.anybill.de/llms.txt and CLAUDE.md, then integrate
the anybill iOS SDK into this project. Use the externalId-based token
flow so we do not have to persist an additional anybill user id.

GitHub Copilot and other tools

  • GitHub Copilot: place the content of AGENTS.md at .github/copilot-instructions.md.
  • Other tools: use the equivalent project-instructions file your tool supports (for example .codeium/instructions.md or tool-specific rules files) and link to https://developer.anybill.de/llms.txt as the documentation source.

Example prompts

Android integration:

Integrate the anybill Android SDK into this existing Kotlin app. The user
is already authenticated via our backend. Use the externalId-based token
flow from the Partner Platform API (POST /v3/user/token with externalId),
reusing our customerId as the externalId so we do not have to persist an
anybill user id. Read the anybill docs at
https://developer.anybill.de/llms.txt for the exact flow.

iOS integration:

Integrate the anybill iOS SDK into this existing Swift app. The user is
already authenticated via our backend. Use the externalId-based token
flow from the Partner Platform API (POST /v3/user/token with externalId),
reusing our customerId as the externalId so we do not have to persist an
anybill user id. Read the anybill docs at
https://developer.anybill.de/llms.txt for the exact flow.

Token refresh callback:

Implement the token refresh callback for the anybill SDK on Android. When
the refresh fails, fetch a new token from our backend endpoint
`/anybill/token` and set it via the SDK's setToken method.

Cross-platform decision:

I need to add digital receipts to both our Android and iOS apps. Compare
the anybill native SDKs vs. the Flutter SDK for our use case, based on the
official anybill documentation.

Security considerations

Keep this short and apply it strictly:

  • Do not let AI agents embed real API keys, partner credentials or production tokens in client code or commit history.
  • Integration tokens for the Frontend SDK are issued by your backend via the Partner Platform API. Minting stays server-side.
  • Review any AI-generated change that touches authentication, token handling or network calls before merging.

See the Security and Compliance page for the broader platform-level guarantees.

Feedback

We actively evolve this setup. If a prompt, a documentation page or the AGENTS.md template did not work as expected, let us know via the usual developer support channel at dev@anybill.de.

Built 2026-09-17 18:08 CEST from commit 3616c61