I got tired of not knowing whether the customer actually got the webhook, so I built a self-hosted gateway (MIT)
retained text
Railhook is a self-hosted webhook gateway. MIT licensed. What it does: sends webhooks to your customers with HMAC signatures, automatic retries and a log of every attempt, so you can see the response body, the status code and when the next retry happens, and replay a delivery manually. It also receives webhooks from Stripe, GitHub, Shopify and 7 other providers, verifies signatures, drops duplicates and forwards them to your service. The CLI forwards live webhooks to localhost while you develop, so no third party tunnel. Also included: an embeddable portal for your own customers, a visual workflow builder for filtering and transforming events, JavaScript transformations. Install: curl -fsSL https://railhook.io/install.sh | bash Docker, about 4 GB RAM. Java, PostgreSQL, Kafka, Redis, all in one compose file. Helm chart available. Built with Claude Code. I designed the architecture, set the project rules and checks, reviewed every change. Failure handling and tests are mine. Demo without signup: https://railhook.io/demo GitHub: https://github.com/vadymkykalo/railhook Feedback welcome, especially on failure handling: restart recovery, duplicate prevention, per endpoint ordering.
document fetched 2026-09-25 00:31 · canonical https://reddit.com/r/selfhosted/comments/1wp05du/i_got_tired_of_not_knowing_whether_the_customer