Configure client-side hooks for automated workflows when AI agents claim and complete tasks.
Important: Hooks are designed exclusively for AI-Optimized boards working with AI agents. They do not execute for regular boards or human users.
Hooks are shell commands that execute on the agent's machine at specific points in the task lifecycle:
git pull)mix test)gh pr create)git push)Create a .stride.md file in your project root with hook definitions. This file is typically created when an AI agent calls the onboarding endpoint and it defines the automation steps that AI agents will execute:
When hooks execute, AI agents receive environment variables with task context:
TASK_ID, TASK_IDENTIFIER, TASK_TITLETASK_STATUS, TASK_COMPLEXITY, TASK_PRIORITYBOARD_NAME, COLUMN_NAME, AGENT_NAMEAll four hooks are blocking - they must succeed for the agent to proceed:
Remember: Hooks are for AI agent automation only. Regular board users won't trigger hook execution.
For comprehensive details on hook execution, including platform-specific examples (Unix/Linux, Windows, macOS), advanced patterns, debugging tips, and best practices, see the complete Agent Hook Execution Guide.
This guide covers:
You've completed all the steps in this guide.