# For AI agents

This docs site follows the [llmstxt.org](https://llmstxt.org) convention so AI agents and automated tools can read the full content without crawling HTML.

## Available resources

- **`/llms.txt`** (at the site root) and **`/docs/llms.txt`** (a copy) — a Markdown link list to every page, in llmstxt.org format.
- **`/llms-full.txt`** (at the site root) and **`/docs/v1/llms-full.txt`** (a copy) — the full Markdown content of every page, concatenated into one file.
- **Parallel Markdown twins** — every HTML page (e.g. `/docs/v1/sdk/stripe/`) has a sibling `index.md` file (`/docs/v1/sdk/stripe/index.md`), linked via `<link rel="alternate" type="text/markdown">` in `<head>` and a "View Markdown" link shown at the top of the page.
- **OpenAPI** — [`openapi.json`](/docs/v1/openapi.json) and [`openapi.yaml`](/docs/v1/openapi.yaml), the full machine-readable specification for every endpoint, parameter, and schema.
- **JWKS** — Ed25519 public keys to verify signed envelopes: `https://keys.vkl-shield.com/v1/keys` (live), `https://keys.vkl-shield.com/sandbox/v1/keys` (sandbox).

## Usage tips

1. Want a quick overview of the site: read `/llms.txt` first — a summary with links to each page.
2. Need the full content in a single call: read `/llms-full.txt`.
3. Need to call the real API: use `openapi.json`/`openapi.yaml` to generate a client or import into Postman/Insomnia.
4. Only need one specific page as plain text (no HTML): append `index.md` to that page's URL.