The first public release of the HookWatch platform. It captures inbound webhooks, forwards them to your destination, and records everything in between so a failed delivery is something you can inspect and re-send rather than guess about.
Added
- Capture and forward — every delivery to an endpoint is recorded end to end:
full request headers and body, the destination’s response status and body, and a
numbered attempt history. Statuses move
received → forwarding → delivered | failed. - Endpoint modes — run an endpoint as a proxy (forward to your destination), capture-only (record without forwarding), or a tunnel that streams to a local process. Ingestion can be paused and resumed per endpoint.
- Automatic retries — opt in per endpoint. Retries fire only on retryable
failures (transport error, timeout, or a 5xx from the destination); 4xx responses
are treated as permanent. The policy covers
max_attemptsandbackoff_secondswith jitter. - Manual replay — re-send the exact captured payload, singly or in a batch. A replay never mutates the original record; it is stored as a separate replay with its own attempt.
- Incidents and alerts — repeated failures on an endpoint are grouped into an incident, and alert rules notify a Slack or outbound-webhook channel on failure spikes, with a log of firings.
- Live dashboard — this release computes endpoint event counts, failure rate, and health over a rolling window so the dashboard reflects current traffic.
See the feature overview for the full picture.
Notes
Release date comes from the v1.0.1 git tag (4 July 2026).