Zyte Api

@Zyte Api what's the current health status of all your service components?

Based on the latest snapshot from Zyte's status page, here's the component health overview: **API Gateway** (operational), **Extraction Service** (operational), **Browser Rendering** (operational), **Data Delivery** (operational), and **Authentication Service** (operational). All 5 core components are currently reporting green status with no degradation. This represents the state as of the last status page update, typically refreshed every 60 seconds.

This example demonstrates the read-heavy component monitoring capability. The MCP queries Zyte's public status API, so no account-specific auth is needed beyond the integration's API key. Status reflects Zyte's self-reported health, not your account's quota or rate limit state.

@Zyte Api are there any unresolved incidents or upcoming maintenance windows I should know about?

Checking Zyte's incident and maintenance feeds: **Unresolved Incidents:** None currently open. **Upcoming Maintenance:** One scheduled window on March 15, 2024, 02:00-04:00 UTC for database infrastructure upgrades. Expected impact: intermittent 503 responses during the window, automatic retry recommended. This maintenance was posted 72 hours in advance per Zyte's SLA. No other events are scheduled in the next 30 days.

This example shows the MCP synthesizing data from two separate tools (unresolved incidents + upcoming maintenance) into a single awareness check. Useful for pre-deployment health checks or daily standups. The MCP cannot subscribe to alerts—this is a point-in-time query only.

@Zyte Api summarize the last 10 incidents by severity and resolution time, then flag any patterns.

Analyzing the 10 most recent incidents from Zyte's status history: **Severity breakdown:** 6 minor (< 5% error rate), 3 major (10-25% error rate), 1 critical (full outage, 18 min duration). **Median resolution time:** 42 minutes. **Pattern detected:** 4 of the 6 minor incidents involved the Browser Rendering component between 18:00-22:00 UTC, suggesting possible load-related degradation during peak hours. The critical incident on Feb 28 was an authentication service failure, resolved via failover. No recurring root causes across major incidents.

This example pairs the MCP's incident-listing tool with the AI's reasoning to surface operational trends. The analysis is illustrative—actual pattern detection depends on incident metadata richness. The MCP fetches the 50 most recent incidents; filtering to 10 and grouping by attributes happens in the AI layer.

When your data pipeline depends on Zyte uptime

A 6-person data team runs nightly scraping jobs that feed a customer analytics dashboard. When Zyte goes down at 3am, the on-call engineer needs to know if it's their code or Zyte's infrastructure before waking the team. This MCP is the right call if Zyte is a single-vendor dependency and you're already using Switchy for runbooks. The 8 status tools give you incident lists, component health, and maintenance windows in one query instead of toggling browser tabs. The trade-off: this is overkill if you only check Zyte status once a quarter or if you're monitoring 12 different SaaS vendors (at that scale, a dedicated status aggregator wins). If Zyte downtime blocks revenue and you need sub-5-minute triage, connect this MCP and write a 'check Zyte health' prompt your on-call can run from Slack.

Avoid deploying into a known maintenance window

A 3-person dev team ships scraper updates every Friday afternoon. Last month they pushed a new extraction rule during an unannounced Zyte maintenance window and spent two hours debugging phantom failures. Now they want a pre-deploy checklist that includes 'is Zyte healthy right now'. This MCP fits if your deploy frequency is weekly or higher and Zyte is mission-critical. The 'Upcoming Scheduled Maintenances' and 'Get Zyte API Status Rollup' tools answer the question in 10 seconds instead of hunting through email or the status page. The boundary: if you deploy once a month, manually checking the status page is faster than setting up the MCP. If you're deploying daily and Zyte is one of five external APIs, this MCP earns its spot in your pre-flight Switchy prompt.

When support needs to rule out vendor issues fast

A 10-person support team fields 'my data is stale' tickets from customers who rely on Zyte-powered feeds. The first question is always 'is this our bug or Zyte's?' but checking the status page mid-ticket breaks flow. This MCP works if support reps are already in Switchy for knowledge lookups and you get more than 3 Zyte-related tickets a week. The 'List Zyte API status incidents' and 'Get Unresolved Incidents' tools let a rep confirm 'yes, Zyte had a 12-minute outage at 9am' without leaving the conversation. The limit: if your support team isn't using Switchy yet, adding this MCP alone won't justify onboarding. If they're already there and Zyte status is a top-5 triage question, wire this into your 'diagnose data delay' prompt and save 90 seconds per ticket.