This article explains the four core building blocks of PulseAPI and how they work together. Understanding these concepts makes every other part of the product easier to use.

Endpoint (Monitor)

An endpoint is a URL you want PulseAPI to watch. In PulseAPI's interface, endpoints are called monitors interchangeably.

When you create a monitor, you configure:

• The URL to request
• The HTTP method (GET, POST, etc.)
• How often to check it (the check interval)
• Optional settings like request headers, a request body, authentication, and an expected response

Each monitor belongs to a project and a team.

Check

A check is a single HTTP request PulseAPI makes to a monitor.

Checks happen automatically on your configured interval (e.g., every 60 seconds). Each check records:

• Status — whether the check was successful or failed, and the specific failure type
• HTTP status code — what the endpoint returned (e.g., 200, 404, 503)
• Response time — how long the request took in milliseconds
• Error message — if the check failed, the reason (e.g., timeout, connection refused, SSL error)

Checks are stored in your check history. How far back history is retained depends on your plan.

Note: A single failed check does not automatically create an incident. The check result is evaluated against your alert rules to determine whether an incident should be created.

Alert Rule

An alert rule defines the condition under which PulseAPI should create an incident and send a notification.

Each rule has:

• A condition type — what to measure: response time, HTTP status code, uptime percentage, or SSL certificate expiry
• An operator and threshold — the specific value that triggers the rule (e.g., response time > 2000ms, status code is 5xx)
• A scope — whether the rule applies to all monitors on your team, or only specific ones
• Notification channels — where to send alerts when the rule fires
• A cooldown period — how long to wait before firing the same rule again, to prevent notification spam
• A priority — which determines the severity level of the resulting incident (critical, high, medium, low)

Rules are evaluated after every check. A rule only fires if the threshold is crossed and the cooldown has elapsed.

Incident

An incident is a record of a problem with one of your monitors. Incidents are created automatically when a check fails and an alert rule condition is met.

Every incident has:

• Status — open, acknowledged, or resolved
• Severity — determined by the alert rule that created it (critical, high, medium, low)
• Affected monitor — which endpoint triggered the incident
• Timeline — when it started, when it was acknowledged, when it was resolved
• Root cause (optional) — a note you can add when resolving

Incident lifecycle:

1. Open — PulseAPI created the incident; your team has been notified
2. Acknowledged — A team member has seen it and is working on it
3. Resolved — The problem is fixed; either auto-resolved when checks pass again, or manually resolved by a team member

Notification Channel

A notification channel is a destination where PulseAPI sends alert notifications. PulseAPI supports:

• Email — send alerts to an email address
• Slack — post alerts to a Slack channel
• Webhook — send a JSON payload to a URL of your choosing

Channels must be added before they can be assigned to alert rules. You can have multiple channels and assign them to different rules independently.

How They Work Together

Monitor → Check (every N seconds)
              ↓
         Alert Rule evaluation
              ↓ (rule matches)
         Incident created
              ↓
         Notification sent to channels
              ↓
         Team acknowledges → resolves

Every 60 seconds (or your configured interval), PulseAPI checks your monitor. The result is evaluated against all active alert rules. If a rule condition is met, an incident is created and your notification channels receive an alert.

Related Articles

• Creating a Monitor
• How Alert Rules Work
• What Is an Incident?
• How Notification Channels Work

Still have questions? Contact support.