CyberNord Relay

Recover failed Stripe webhooks safely in minutes.

See the reason, inspect every delivery attempt, diff payload changes, replay safely and recover payment or platform webhooks before customers notice.

First successful webhook in 3 minutes: 1. Apply a Stripe, Shopify, GitHub or GitLab template. 2. Create the endpoint. 3. Send one test event to the generated /in/{token} URL. 4. Inspect diagnostics. 5. Turn on alerts before going live.

Endpoints

0

Events Today

0

Daily usage meter.

Failed Events

0

Success Rate

100%

Fast onboarding

Zero-config start for Stripe, Shopify, GitHub and GitLab

Pick a template, keep forwarding empty for inspect-only mode or point it to your consumer, then send one test event and confirm delivery in the feed.

Onboarding

First healthy event

Apply template and keep validation on.

Create endpoint and copy the generated ingress URL.

Send one smoke payload and inspect event status, signature and latency.

Payments wedge

Stripe reliability

Retryable `429`, `5xx` and timeout delivery states.

Payment-focused diagnostics and replay controls.

Timeline that shows every delivery attempt.

Upgrade path

Plan loading…

Checking current plan and entitlements.

Why teams keep Relay open

“We stopped digging through target logs just to answer whether Stripe even reached us.”Payments team · beta

“The replay flow gave support a safe path to recover missed billing events without waking backend engineers.”SaaS operations · internal pilot

“Incident grouping turned fifty red rows into one problem we could assign, explain and close.”Platform team · CyberNord demo

Workspace

Loading workspace…

Preparing CyberNord product context.

Create endpoint

Forwarding target URL example: `http://host.docker.internal:9090/webhook` for a local consumer in Docker, or `https://api.your-service.com/webhooks/incoming` for production.

Create recommended alert rules Tip: apply a provider template to unlock rule templates (payments, signature, CI, gate).

Zero-config path: use a provider template first, keep validation enabled for signed providers, and leave forwarding empty if you only want to inspect ingress before wiring production delivery.

Existing endpoints

Endpoint	Environment	Events	Success	Last Event

Endpoint details

Select an endpoint

Tabs below open analytics, grouped incidents, replay history, settings and members for the selected endpoint.

No endpoint selected yet.

Quarantine

Dead Letter Queue (DLQ)

Terminal failures are quarantined here for safe recovery. Group by root cause, batch replay into staging, or discard/resume with an audit trail.

DLQ: loading…

Groups

Grouped by reason + provider + event type + status-code family.

Items

Select a group to inspect items and run batch actions.

Batch replay uses your existing replay engine and keeps per-event auditability.

Flows

Satellite mode lite

Top correlation chains across endpoints, with the first failure point and a mini multi-hop story. This is the root-cause screen.

Provider templates

Popular and custom presets

Alert rules

Slack, Telegram, Discord and generic webhooks

Cooldown presets Advanced JSON

Telegram target: bot_token=123:abc&chat_id=987654321. Slack/Discord: raw incoming webhook URL. GitHub issue: repo=owner/name&token=ghp_xxx (optional labels=a,b). Jira issue: base_url=https://acme.atlassian.net&email=...&api_token=...&project=KEY&issue_type=Bug. Threshold examples: {"count":3,"window_minutes":10}, {"latency_ms":3000}, {"success_rate":0.95}.

Pick a preset or rule type and CyberNord Relay will suggest a good default.

Triggered alerts

Recent incidents

Live Live: off

Tip: signal:ci, signal:deploy, ip:1.2.3.4

Advanced filters

Signature issues Replayed Incident-worthy only

Use search above to find failed events, then paste selected IDs here. Alternate target is useful for replaying into staging before you hit production again.

DSL examples: status:failed provider:stripe latency>3000, provider:github signature:failed attempts>1, incident:true retryable:true tag:timeout, provider_event_id:evt_123 confidence>0.8, after:2026-03-20 before:2026-03-24.

Time	Endpoint	Provider	Status	Method	Latency	Flags

Showing latest events.

Diagnostics

Selected event

Inspect payload, response, signature state and recommended recovery actions.

No event selected yet.

Smart replay

Edit and resend

Override headers, body, alternate target and delay without losing auditability.

Queue simulation

Leave target empty to use the endpoint default. Fallback target is used for retry/failover scenarios. Use delay when you want to replay after fixing downstream infra. Headers JSON must be an object of header names to string arrays, for example `{"X-Test":["1"]}`.

Provider setup wizard

Stripe, GitHub and GitLab step-by-step

Use this when you want exact provider wiring, header names, target hints and a smoke-test flow without jumping between docs.

Setup preview

Pick a provider to see exact setup steps.

Integration lab

One-click test payloads for Stripe, GitHub, GitLab and OpenAI

Load a realistic sample, send it to the selected endpoint, then inspect diagnostics, tags, retries and incidents without leaving the product.

For signed providers, the lab uses the helper secrets/tokens from the Provider helpers block below. Select an endpoint first so the lab knows where to send the sample.

Lab output

Pick a provider, load a sample, then send it to the selected endpoint.

Provider helpers

Stripe, Shopify, GitHub, GitLab and OpenAI setup

Provider-specific setup helpers make smoke tests and validation less guessy.

Stripe helper

Paste a header to confirm whether the secret, timestamp and payload match.

Stripe curl preview will appear here

Shopify helper

Shopify curl preview will appear here

GitLab helper

GitLab curl preview will appear here

GitHub helper

GitHub curl preview will appear here

Local proxy

Forward cloud webhooks to localhost in one command

Streams events from CyberNord Relay and forwards them to your local service (no ngrok).

Open docs →

1) Select endpoint

Pick an endpoint in Dashboard. The command below uses the selected endpoint ID.

Works best for local dev when your consumer can accept the same path as the provider.

2) Run tunnel

Export CN_RELAY_TOKEN and run cyber-relay dev.

Use -dsl or -search to filter traffic.

Tunnel settings

Local port

Target base URL

Path mode

Tunnel presets

Presets auto-fill DSL and clear overly strict filters.

Troubleshooting

Loading…

Focus

Focus buttons update the DSL (provider + mode) so you can quickly narrow the stream.

Stream filter (DSL)

Stream search

Change values above to update the command in real-time.

Command

Loading…

Magic one-liner (fast dev)

Runs the tunnel without selecting an endpoint (auto-picks). Great for first-time setup and demos.

Loading…

Tip

For safe recovery: inspect event → fix target/secret → use Auto-fix + replay (staging first).

On-prem

Satellite fleet

Manage on-prem tunnel agents (“satellites”) running inside private networks. Satellites send heartbeats so you can see health, versions and last seen IP without opening inbound ports.

Create satellite token

Token is shown once. Store it in a secret manager and pass it to the agent as CN_SATELLITE_TOKEN.

Fleet

Loading…

Usage & limits

Loading usage…

Features

Loading entitlements…

PII / GDPR masking

Mask sensitive data in UI + API responses by default. Storage remains raw for replay/forensics (enterprise option: mask-at-ingest will be added in Phase 3).

Enable masking

RBAC policy (fine-grained)

Optional overrides for minimum roles per action. Format: {"replay":"editor","dlq_batch":"editor","members_manage":"admin"}. Roles: viewer/editor/admin/owner.

Assistant

Setup and troubleshooting assistant

Core diagnostics always work without remote AI. Pro and above unlock optional AI modes such as local LLM, BYOK, and managed CyberNord AI when available.

Use your OpenAI API key (BYOK)

Personal OpenAI API key

Use local/free LLM (Ollama / LM Studio)

Advanced: CyberNord OpenAI (paid plans)

Keep this off if you want a zero-cost assistant. Use Local LLM above or BYOK.

Enable CyberNord OpenAI

Free and Founder Access keep built-in diagnostics only. Remote AI modes unlock on Pro and above.

Product release version

Managed via CyberNord Product CMS and dashboard release control.

Current version: —

Open Product CMS

Release markers

Add deploy markers to overlay charts and unlock “after deploy it got worse” insights.

CI failed notifier + deploy hooks

Create the CI / Deploy hooks endpoint template, then post pipeline/deploy results to /ci/in/{token}. CI failures become events + alerts. Successful deploy signals can auto-create release markers for deploy impact insights.

Example (GitHub Actions / GitLab CI):

curl -sS -X POST "https://<YOUR-RELAY-DOMAIN>/ci/in/<TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{"kind":"deploy","status":"success","version":"v1.2.3","environment":"prod","provider":"github_actions","commit_sha":"abc123","branch":"main","url":"https://github.com/org/repo/actions/runs/123"}'

Use status: failed to trigger your any_failure alert rules. Use kind: deploy + status: success + version to auto-create release markers.

Checking OpenAI assistant health…

Assistant answers and built-in docs will appear here.