Collection Runs

Postman groups API requests into collections that can be executed together. An agent runs a named collection on a trigger, reads each response and the pass or fail of its tests, and applies your rule, for example opening a ticket when an assertion breaks. It records the run outcome and notifies the owning team. Runs where many requests fail at once, or results that differ sharply from the last known good run, are escalated to an engineer rather than filed automatically, so a broken dependency gets a person's attention before it spreads downstream.

Environment & Variable Records

Postman environments hold the variables, base URLs, and tokens that requests depend on. An agent reads the environment tied to a run, applies your rule to confirm required values are present, and updates a variable where you have cleared it to do so. It notifies the owner when something expected is missing. Secrets that look expired, or variables the agent is not permitted to change, are left untouched and handed to a person, so credentials and configuration stay under human control instead of being rewritten by an automated step.

Monitors & Alerts

Postman monitors run collections on a schedule and record uptime and response health over time. An agent reads monitor results, applies your rule, and, for instance, marks a service healthy or raises an incident when checks fail repeatedly. It updates the status record and messages on-call. A single flaky failure is observed rather than escalated, while sustained or widespread failures are routed to a person, so alerts reflect real problems and the team is not paged for noise the agent could reasonably absorb.