Loop icon
Download icon
Get icon
Script icon
If icon
ChatCompletion icon
GoogleGemini icon
SlackIncomingWebhook icon
Set icon
Schedule icon
Webhook icon

Watch SaaS Vendor Status Pages and Get an AI Impact Assessment in Slack

Detect new incidents on GitHub, Cloudflare, Stripe and other status pages and post a Slack alert with an AI assessment of the impact on your product.

Categories
AIInfrastructure

When GitHub, Cloudflare, Stripe or your LLM provider has an incident, the first question on call is "does this affect us, and what should we do?". Status page emails do not answer that, and they arrive for every minor component. This blueprint polls the public status pages of the vendors you depend on, detects incidents that are new or have a new update, and posts a Slack alert with an AI assessment based on how you use each vendor.

It works with any status page hosted on Atlassian Statuspage, which exposes the public /api/v2/incidents/unresolved.json endpoint. Many vendors use it, including GitHub, Cloudflare, Anthropic, Vercel, Atlassian and Stripe.

How it works

For each vendor in vendors (Loop, three at a time):

  1. fetch_incidents downloads the vendor's unresolved incidents.
  2. load_known reads the incident updates already reported, from the Kestra KV store.
  3. find_new_updates keeps incidents that are new or have a newer update than last time.
  4. if_fresh skips the LLM and Slack when nothing changed.
  5. assess_impact sends the incident text, affected components and your usage description to Gemini and gets structured JSON back: severity_for_us, impact and action.
  6. alert_slack posts the incident facts (taken from the status page, never from the model) and the assessment.
  7. remember_updates stores what was reported, after the alert. Resolved incidents drop out of the feed, so the stored state stays small.

The errors branch posts to Slack if a status page cannot be read.

Prerequisites

  • A Google AI Studio API key for Gemini (the free tier covers typical incident volumes).
  • A Slack Incoming Webhook.

Secrets

  • GEMINI_API_KEY: Gemini API key used by assess_impact.
  • SLACK_WEBHOOK_URL: Slack Incoming Webhook URL for alerts and failures.

In the open-source edition, secrets are environment variables on the Kestra container, prefixed with SECRET_ and base64-encoded. See how to manage secrets.

Inputs

  • vendors (JSON): list of {"key", "name", "status_url", "usage"} objects. status_url is the status page base URL (for example https://www.githubstatus.com). usage is how you depend on the vendor, which drives the assessment. key names the KV entry status_watch_<key>.
  • product_context (STRING): what your product is and who uses it.
  • model_name (STRING, default gemini-2.5-flash).

Quick start

  1. Add the two secrets and edit vendors and product_context.
  2. Execute the flow. Every currently unresolved incident is reported once.
  3. Execute it again: only incidents with a new update are reported.
  4. Enable every_five_minutes. Change the webhook key before using on_demand.

To find a vendor's endpoint, open its status page and append /api/v2/incidents/unresolved.json. If that returns JSON, the vendor is supported.

Outputs

  • {{ outputs.find_new_updates.vars.fresh }}: new or updated incidents per vendor.
  • {{ outputs.assess_impact.jsonOutput }}: the AI assessment per vendor with fresh incidents.
  • KV entries status_watch_<key>: the latest reported update per incident.

Links

See How

New to Kestra?

Use blueprints to kickstart your first workflows.