Using OpenReceive with coding agents
OpenReceive publishes machine-readable specifications and installable agent skills, so a coding agent can add Bitcoin Lightning checkout to your application — or debug one — without guessing.
Building an integration?
Install the integrate-openreceive skill (it detects your stack and carries the full quickstart), or paste the one-prompt agent directions:
- Claude Code:
/plugin marketplace add OpenReceive/openreceive, then/plugin install openreceive - Skills CLI (Codex, Cursor, and other SKILL.md-compatible tools):
npx skills add OpenReceive/openreceive - For GitHub Copilot and other agents, install into your application’s
.agents/skills/withnpx skills add OpenReceive/openreceiveor the package command below. The OpenReceive repository’s own.agents/skills/is only discovered when that repository is the workspace. - No installer? Copy the agent directions for your stack and paste them into
your agent: Node (Express) ·
Fastify ·
Next.js ·
Rails ·
PHP ·
Laravel ·
Django ·
FastAPI ·
WordPress + WooCommerce ·
BTCPay Server. Each
is self-contained, quickstart included. To hand an agent a link instead, drop
/fullfrom the URL: that short page tells it to download the full file with its shell.
One package per ecosystem bundles an offline copy: @openreceive/node, the
core openreceive gem, Python’s openreceive, and Composer’s
openreceive/openreceive. From your application’s directory, install that copy
where your agent can discover it:
| Stack | Install bundled skills |
|---|---|
| Node | npx openreceive skills install |
| Python (Django or FastAPI) | openreceive skills install |
| Rails | bin/rails openreceive:skills |
| Laravel | php artisan openreceive:skills |
These commands write both skills to .agents/skills/; add
--dir .claude/skills for Claude Code. Re-running replaces only the two
OpenReceive skill folders, removing obsolete files and leaving other skills
alone. Plain PHP, non-Rails Ruby, and WordPress projects can use
npx skills add OpenReceive/openreceive. Installing a library alone does not
make its bundled skills discoverable.
A debug-openreceive-payment skill ships alongside: boot failures, 403/404/ 409 semantics, settlement timing, swap refunds, each with its fix.
Working on OpenReceive itself?
Read AGENTS.md in the repository — the architectural invariants a change must not violate.
Generating an HTTP client, or verifying routes?
Use the normative OpenAPI contract:
https://openreceive.org/openapi.yaml.
It is the exact spec/openapi/openreceive-http.v1.yaml file from the
repository — the same contract the shipped adapters and the Rails engine are
tested against.
Giving an agent documentation context?
Start from https://openreceive.org/llms.txt
— every guide as raw markdown, one fetch away. Any guide page is also
available as markdown by appending .md to its URL.
Questions, or a problem with the library itself: https://openreceive.org/contact