Push by Techulus

@Push send a notification to my mobile devices: 'Deployment to production completed successfully. Build #247 is now live.'

Notification sent asynchronously to all devices linked to your Push account. The message 'Deployment to production completed successfully. Build #247 is now live.' was delivered with default sound and priority settings. Delivery typically completes within 2-3 seconds. You can check the Push dashboard for per-device delivery confirmation and read receipts if your devices have that enabled.

This example uses the async send tool, which Push recommends for accounts with 10+ devices. The MCP requires a valid API key with notification-send permissions. Note that delivery confirmation is eventual — the tool returns immediately but actual device receipt depends on network conditions and device state (locked, do-not-disturb, etc.).

@Push invite sarah.chen@example.com to our team so she can receive deployment alerts

Invitation sent to sarah.chen@example.com. An email with a team join link was dispatched through Push's system. Once Sarah accepts the invitation, she'll be able to link her devices to the team and receive notifications sent to the team's API key. The invite remains valid for 7 days. You can revoke it before acceptance using the team management tools.

This demonstrates team collaboration features. The invitee must have a Push account (or create one) to accept. Once joined, any notification sent to the team's API key will reach all members' devices. Useful for on-call rotations or shared alerting, but be aware that team members see all notifications — there's no per-user filtering at the MCP level.

@Push check which team members are currently active, then send a notification only to the 'on-call' device group about the database latency spike we're seeing

The MCP doesn't expose a tool to list team members or query device groups, so I can't retrieve the active roster. However, I can send a notification to your 'on-call' device group if you've pre-configured it in the Push dashboard. Sending now: 'Database latency spike detected. Current p95: 450ms (threshold: 200ms). Investigate connection pool saturation.' This targets only devices tagged with the 'on-call' group label, avoiding alert fatigue for off-duty team members.

This example highlights a gap: the MCP provides send and invite tools but no read operations for team state or device inventory. You must manage device groups through Push's web UI. The device-group send tool is powerful for targeted alerting, but group membership is opaque to the AI — you'll need to tell it which groups exist and what they're for.

When Push wins for waking up the right engineer at 3am

A 6-person SaaS ops team runs a PagerDuty-style rotation but needs custom escalation logic that their monitoring stack doesn't support. Push's device-group targeting lets them send critical alerts to the on-call engineer's phone and tablet simultaneously, with fallback to the secondary if no ack in 2 minutes. The async send tool handles bursts during multi-service outages without blocking the alert pipeline. The trade-off: if your escalation logic is simple (one person, one device), PagerDuty or Opsgenie is overkill but still cheaper than building custom routing. Push makes sense when you need programmable notification delivery that your existing tools can't express, and you're comfortable managing API keys per team or rotation slot.

Push works when Slack channels aren't enough for triage

A 12-person support team uses Zendesk but needs instant mobile alerts for VIP tickets that bypass the shared Slack channel noise. Push's webhook tool lets them POST from Zendesk triggers directly to the VIP-support device group, so the 3 senior reps get push notifications on their phones within seconds of ticket creation. The team-invite and delete-member tools keep the device group in sync as reps rotate on/off VIP duty each week. The boundary: if your team is under 5 people or everyone monitors the same queue, a Slack channel with @mentions is simpler and free. Push pays off when you need selective routing to subsets of a larger team and mobile delivery matters more than chat-thread context.

When Push beats email for breaking the right developer's flow

A 4-person product team runs CI/CD on GitHub Actions but loses build failures in email noise and Slack's scroll-back. Push's sync notification tool sends a blocking alert to the commit author's phone the moment their PR breaks main, with a deep link to the failed job. The API-key auth means each developer registers their own devices without sharing credentials, and the team admin uses the invite tool to onboard new hires in under a minute. The limit: if your team is disciplined about checking Slack or your build times are under 2 minutes, the interruption cost outweighs the speed gain. Push is the right call when build feedback loops are long enough that context-switching to fix them is cheaper than waiting for the next Slack check.