Status Page Incident Digest Monitor

Monitor public vendor status pages and incident RSS feeds. Returns one executive digest row per vendor with change detection, incident state classification, affected components, and actionable recommendations.

Quickstart (1 minute)

1. Click Try for free and paste this input:

{
  "vendors": [
    {
      "id": "stripe",
      "name": "Stripe",
      "criticality": "high",
      "owner": "Ops",
      "statusPageUrl": "https://status.stripe.com",
      "incidentFeedUrl": "https://status.stripe.com/history.rss"
    }
  ],
  "delivery": "dataset",
  "datasetMode": "all",
  "snapshotKey": "status-page-quickstart"
}

2. First run seeds the baseline snapshot.
3. Second run shows changedSinceLastRun for any vendors that changed state.

Output Shape

Each row in the dataset represents one vendor:

Field	Type	Description
vendorId	string	Stable slug identifier
vendorName	string	Human-readable name
executiveSummary	string	One-liner plain-English status
incidentState	enum	operational / degraded / outage / maintenance / unknown
incidentCount	integer	Active / recent incidents in the lookback window
affectedComponents	string[]	Component names that are not fully operational
actionNeeded	boolean	true when outage, degraded, or state changed for high-criticality
recommendedActions	string[]	Actionable steps for the on-call owner
changedSinceLastRun	boolean	State or components changed vs previous snapshot
evidence	object[]	Raw incidents / feed items supporting the digest
meta	object	Vendor config + parse diagnostics

Input Reference

Vendor Object

Field	Required	Description
id	No	Stable slug (auto-generated from name if omitted)
name	No	Display name
statusPageUrl	Yes	Public status page URL
incidentFeedUrl	No	RSS / Atom incident feed URL
criticality	No	mission_critical / high / standard / low (default: standard)
owner	No	Team responsible for this vendor
tags	No	Freeform tags for filtering

Top-Level Options

Field	Default	Description
delivery	dataset	dataset or webhook
datasetMode	all	all / changes_only / action_needed
webhookUrl	—	Required when delivery=webhook
notifyOnNoChange	false	Fire webhook even when nothing changed
snapshotKey	status-page-incident-snapshots	Key for run-to-run state
incidentLookbackDays	30	Days of feed history to include (0 = all)
maxIncidentsPerVendor	10	Max evidence rows per vendor
concurrency	3	Parallel vendor fetches
requestTimeoutSeconds	30	HTTP timeout per request
dryRun	false	Preview without saving snapshots

Supported Status Page Formats

1. statuspage.io JSON API – Automatically tries /api/v2/summary.json for any status page URL. Returns structured component status, active incidents, and scheduled maintenance.
2. HTML scraping fallback – Keyword-based detection of operational / degraded / outage / maintenance state from any public status page HTML.
3. RSS / Atom incident feed – Parses incident history feed; respects incidentLookbackDays.

Templates

Template	Use Case
quickstart_dataset	Single vendor, all rows, baseline first run
recurring_changes_only	Daily schedule, multi-vendor, changes only
webhook_action_needed	Webhook alert when actionNeeded=true

Change Detection

State is fingerprinted as a SHA-256 hash of (incidentState, sortedAffectedComponents, incidentCount). On each run the fingerprint is compared to the stored snapshot. changedSinceLastRun=true means something in that tuple changed.

The first run always sets isFirstRun=true and changedSinceLastRun=false (no previous state to compare).

FAQ

Q: Which vendors work without an incidentFeedUrl?
A: All vendors. The feed is optional and supplements the status page parsing.

Q: Does this use a headless browser?
A: No. Pure HTTP + JSON/RSS parsing. Low resource usage, high throughput.

Q: How do I monitor more than 3 vendors?
A: Add more vendor objects to the vendors array. Use concurrency: 5 for faster runs.

Q: What happens if a status page is unreachable?
A: The vendor row gets incidentState: "unknown" and actionNeeded: true with an error explanation in recommendedActions.