Security Incident Response (Sentry + Datadog + Slack + Linear)
Reduce mean-time-to-diagnose by chaining error tracking + metrics + ticketing into a single Claude-driven workflow.
MCPs in this workflow
Sentry MCP
Query new errors + frequency
Datadog MCP
Cross-check with metric anomalies
GitHub MCP
Find recent deploys / commits
Slack MCP
Post diagnosis to #oncall
Linear MCP
Open incident ticket
Steps
- 01Sentry webhook fires on error spike → triggers Claude
- 02Claude queries error details + same time-window metrics
- 03Cross-refs with recent deploys to suggest probable cause
- 04Posts structured diagnosis to Slack #oncall
- 05Opens Linear incident with severity / labels
▸ outcome
On-call wakes up to diagnosis ready, not "something is wrong".
Build it yourself
All MCPs in this recipe are in the Unyly catalog. Open and install the ones you need in a couple clicks.
Open catalogMore workflows
▸ workflow
Automate Pull Request Review with Claude (GitHub + Slack + Linear)
▸ workflow
Auto-Generate Daily Standup (GitHub + Linear + Slack)
▸ workflow
Auto-Triage Email Inbox (Gmail + Linear + Slack)
▸ workflow
Auto-Prep for Meetings (Calendar + Notion + GitHub)
▸ workflow
Customer Support Triage (Gmail + Notion + Slack)
▸ workflow
Blog Content Pipeline (Notion + GitHub + Twitter)