dev

To route development requests to the correct Cortex workflow, analyze the input to delegate tasks to lifecycle, overnight, or discovery skills. It triages the backlog when no specific feature is named.

When no specific feature is requested, backlog triage handles the items by coordinating with the cortex/backlog directory. This ensures unassigned development tasks are properly routed.

Yes, you need a populated cortex/backlog directory to run Cortex workflows effectively. The routing process relies on this directory to triage requests when no explicit feature is named.

No, you cannot execute Cortex routing from a subdirectory because the system enforces project-root execution. This ensures all prerequisite paths and directories exist before routing requests.

The best way to triage a development backlog using Cortex is to submit a request without naming a specific feature. The routing hub automatically analyzes the input and delegates it to the appropriate workflow.

Cortex workflow routing fails to find prerequisite directories if it is not executed from the project root. The system checks for a populated cortex/backlog directory before routing.