Triage and remediate DevOps alerts with Claude, ClickUp, GitHub and Slack
Go to Workflow
Oneclick AI Squad
Description
Quick overview
This workflow receives monitoring alerts via webhook, uses Anthropic Claude to triage and draft remediation plans, pulls runbooks and related context from a documentation API, GitHub, ClickUp, and Terraform Cloud, then either executes allowlisted low-risk fixes or escalates to on-call via ClickUp and Slack.
How it works
Receives a monitoring alert payload through a POST webhook.
Uses Anthropic Claude to classify the alert and determine whether it represents a critical production outage.
If the alert is critical, creates an urgent ClickUp incident ticket and pages the on-call engineer in Slack, bypassing any automated remediation.
For non-critical alerts, retrieves matching runbooks from a documentation knowledge base API, pulls related GitHub issues/PRs, fetches recent ClickUp incident history, and reads the current Terraform Cloud workspace state.
Uses Anthropic Claude to generate a remediation plan grounded in the retrieved context and extracts the proposed action type, risk level, confidence, and command.
If the plan is allowlisted, low-risk, and high-confidence, calls an HTTP remediation executor to run the fix, logs a resolved ClickUp ticket, and posts confirmation to a DevOps Slack channel.
Otherwise, opens a ClickUp ticket for manual remediation and notifies on-call in Slack with the plan and supporting context.
Logs the alert details, plan, and outcome to an incident knowledge log API for future reference.
Setup
Create and configure the incoming alert source to send POST requests to the workflow webhook URL (path: /devops-alert-received).
Add credentials for Anthropic Claude, GitHub, ClickUp, Slack, and HTTP header auth credentials for your runbook knowledge base API, Terraform Cloud/Enterprise API, remediation executor API, and incident knowledge log API.
Update the configuration values for GitHub owner/repo, ClickUp list ID, Terraform workspace ID, and the runbook, remediation executor, and incident log API URLs.
Set the Slack channel names/IDs for critical paging, on-call notifications, and DevOps automation updates.
Review and tailor the auto-remediation allowlist (action types), along with the risk and confidence thresholds used to permit automatic execution.