Operations and growth teams often want the same weekly numbers in Slack without opening another dashboard. Trends MCP exposes get_trends, get_growth, and get_top_trends through a single POST contract so a small relay service can format blocks when thresholds trip.
Slack is where approvals, launches, and incident threads already live. A short message that cites the same JSON fields each week is easier to audit than a screenshot chain. Trends MCP returns structured arrays for historical pulls and ranked pairs for live feeds, which maps cleanly to Slack Block Kit sections when a relay formats the payload.
The relay can be a Cloud Function, a small Fly.io service, or a worker in an existing API. It should hold the Trends MCP API key, enforce allowlists for keywords, and normalize errors before Slack sees them. The official docs describe error shapes such as missing_parameter, rate_limited, and data_unavailable, which the relay can translate into operator language without leaking secrets.
get_growth responses include period, growth, direction, recent_date, and baseline_date per window. Those five strings are enough for a headline and a context line. When volume_available is true, recent_volume and baseline_volume support a second sentence for stakeholders who still think in absolute counts.
get_top_trends responses include as_of_ts plus ranked rows for the requested type string. Slack messages should print the feed label exactly as documented so humans can reproduce the call during reviews.
One pattern that survives busy channels is three lines: primary metric, window, and source. Example layout: twelve month Google Search growth for the brand term, the numeric percent, and a link to the internal runbook. A second attachment can list Amazon or TikTok growth for the same phrase when commerce teams share the channel.
MCP inside Cursor or Claude is built for interactive reasoning. Slack alerts are scheduled or event driven. Both can share one Trends MCP key organizationally, but the alerting path should stay on REST POST to https://api.trendsmcp.ai/api so cron systems do not depend on an MCP session.
Each successful get_trends or get_growth call counts toward the plan. Pagination for get_top_trends counts per feed and per page as described in the docs. Teams that run hourly Slack tests burn quota fast. Weekly growth summaries plus one daily leaderboard fetch is a common compromise for marketing war rooms.
No-code bridges remain valid when engineering time is scarce. Zapier trend workflows and n8n plus Trends MCP pages describe webhook-first patterns. A custom relay wins when policies require private networking, signature verification on inbound Slack events, or custom deduplication keyed to product lines.
Real-time trends API explains leaderboards and freshness semantics. Brand monitoring ties search and news sentiment series to comms playbooks. GitHub Actions trend reports shows the same JSON contract inside CI when engineering wants parity with Slack.
Normalized scores are comparable within a source over time, but they are still model outputs from upstream pipelines. Empty weeks or data_unavailable responses happen. Good Slack copy mentions the pull timestamp and tells readers which primary source produced the number, especially when TikTok hashtag volume sits next to Google Search growth in one thread.
FAQ