Skip to main content

Slack Integration for Runbooks

Last updated on

Harness AI SRE integrates with Slack at the organization level, enabling automated incident communication and team collaboration across all projects.

Overview​

Slack integration enables your runbooks to:

  • Send automated notifications
  • Create incident-specific channels
  • Manage threaded discussions
  • Coordinate response teams
  • Track incident updates

Integration Setup​

Prerequisites​

  • Slack Workspace Admin access
  • Harness Organization Admin role

Setup Steps​

  1. Navigate to Organization Settings → Third Party Integrations (AI SRE)
  2. Click Connect for Slack
  3. Follow the OAuth flow to authorize Harness
  4. Configure workspace permissions

Required Slack Permissions​

The Harness Slack bot requires these permissions:

  • channels:manage - Create and manage channels
  • chat:write - Send messages
  • groups:write - Manage private channels
  • im:write - Send direct messages

Features​

  • Global Slack workspace access across all projects
  • Unified authentication
  • Centralized channel management
  • Cross-project notifications

Using Slack Actions in Runbooks​

When you add Slack actions to a runbook, you'll configure them through a form-based interface. The specific fields depend on the action type you select.

Send Slack Message Action​

Sends a message to a specified Slack channel.

Form Fields:

  • Channel: Channel name or ID (e.g., #incidents or {{Activity.slack_channel}})
  • Message: Message text to send
    • Supports Mustache variables: {{Activity.title}}, {{Activity.summary}}
    • Can include Slack formatting (bold, italics, links)

Create Slack Channel Action​

Creates a new Slack channel for incident coordination.

Form Fields:

  • Channel Name: Name for the new channel (must follow Slack naming rules)
    • Example: incident-{{Activity.id}}
  • Description: Channel topic/description
  • Is Private: Whether to create a private channel

Available Mustache Variables:

  • {{Activity.title}} - AI SRE incident title
  • {{Activity.id}} - AI SRE incident ID
  • {{Activity.severity}} - AI SRE incident severity
  • {{Activity.status}} - AI SRE incident status
  • Any custom incident fields configured in your incident template

Best Practices​

Channel Naming​

  • Use consistent prefixes
  • Include incident IDs
  • Keep names descriptive
  • Follow workspace conventions

Message Structure​

  • Use clear formatting
  • Include severity indicators
  • Link to relevant dashboards
  • Mention appropriate teams

Permissions​

  • Use least privilege access
  • Regularly audit permissions
  • Document access requirements
  • Monitor usage patterns

Common Use Cases​

Incident Coordination​

  1. Create incident channel
  2. Notify stakeholders
  3. Share initial assessment
  4. Track response actions

Status Updates​

  1. Send periodic updates
  2. Track resolution progress
  3. Share metrics and graphs
  4. Document action items

Post-Incident​

  1. Archive incident channel
  2. Share incident summary
  3. Schedule retrospective
  4. Document lessons learned

Troubleshooting​

Common Issues​

  1. Authentication Failures

    • Verify OAuth tokens
    • Check permission scopes
    • Confirm workspace access

  2. Channel Creation Errors

    • Check naming conventions
    • Verify channel limits
    • Confirm bot permissions

  3. Message Failures

    • Validate channel existence
    • Check message formatting
    • Verify rate limits

Next Steps​