Automations in Palmier allow you to automatically trigger agents based on events from your integrated platforms. When specific conditions are met, your chosen agent will run automatically, helping you maintain consistent workflows and respond to changes in real-time.

​
How Automations Work

Automations are built around three core components:
  1. Trigger: The event that starts the automation (e.g., pull request opened, issue labeled)
  2. Conditions: Optional filters to refine when the automation runs
  3. Agent: The Palmier agent that executes when triggered

​
Available Integrations

Automations are powered by your connected integrations. Each integration provides different triggers and capabilities:

GitHub Automations

Pull requests, issues, and workflow runs

Slack Automations

Coming Soon: Messages, channels, and notifications

Time-Based Automations

Coming Soon: Scheduled runs and periodic checks

​
Need a Different Integration?

We’re continuously expanding our integration ecosystem. If you need support for a specific platform, let us know!

Request Integration

Contact us to prioritize the integrations you need most

​
Creating Automations

To create a new automation:
  1. Navigate to Automations: Go to your project dashboard and click on “Automations”
  2. Choose Integration: Select the integration platform (GitHub, Slack, etc.)
  3. Select Repositories: Choose one or multiple repositories where the automation should be active
  4. Configure Trigger: Choose the event that will trigger your automation
  5. Set Conditions: Add optional filters to refine when the automation runs
  6. Select Additional Repositories: Optionally choose additional repositories that the agent should have access to during execution, beyond the primary trigger repository
  7. Add Instructions: Optionally provide additional context or instructions for the agent (e.g., documentation location, specific requirements)
  8. Select Agent: Choose which agent should execute when triggered
  9. Test & Deploy: Save your automation and test it with real events
Automations require active integrations to function. Make sure you’ve connected and configured your desired integration before creating automations.

​
Multi-Repository Access

When configuring automations, you can grant agents access to additional repositories beyond the primary trigger repository. This enables agents to work across multiple codebases during a single automation run. Use Cases:
  • Cross-repository updates: Update documentation in a docs repository when code changes in the main repository
  • Monorepo coordination: Coordinate changes across multiple related repositories
  • Dependency management: Update dependent repositories when core libraries change
  • Comprehensive analysis: Analyze related codebases together for better context
The agent will have read and write access to all selected repositories during execution, allowing for comprehensive automation workflows that span multiple projects. A run triggered by an automation will have an “Auto” tag on the main page. Hover over it to view which automation triggered that run.
You can specify multiple agents for a single trigger, which will create separate independent runs for each agent. This allows you to have different agents work on the same event simultaneously.

​
Automation Response Behavior

When an automation triggers a run, the agent will automatically respond back to the original trigger location upon completion or when requesting user input. Example Workflow:
  1. Trigger: Pull request opened → Review agent automation activates
  2. Execution: Agent performs code review in its sandbox environment
  3. Response: Agent posts review comments directly on the triggering pull request
  4. Follow-up: Any subsequent interactions within the Palmier web app remain internal and won’t post back to the original PR
This ensures seamless integration between your automation workflows and your existing development platforms, while keeping follow-up discussions contained within Palmier when appropriate.

​
Best Practices

  • Start Simple: Begin with basic triggers and add complexity as needed
  • Test Thoroughly: Use test repositories or channels to verify automation behavior
  • Monitor Performance: Check automation logs to ensure they’re running as expected
  • Add Descriptions: Add detailed descriptions to identify specific automations
You can have multiple automations for the same trigger with different conditions, allowing for complex workflow orchestration.