ai-support

# Support ## Purpose Structured customer support investigation. Organizes findings by ticket, links to relevant code and PRs, and builds a searchable knowledge base of resolved issues. ## Trigger - Command: `/ai-support start <ticket-id>` or `/ai-support find [query]` - Context: customer-reported issue, support ticket investigation, escalation from support team. ## When to Use - Investigating a customer-reported bug or behavior - Reproducing an issue from a support ticket - Documenting resolution for future reference - Escalation requiring code-level investigation ## When NOT to Use - **Production incidents** -- use `/ai-postmortem` - **Internal bugs found during development** -- use `/ai-debug` - **Feature requests** -- create a GitHub Issue with the `enhancement` label ## Modes ### start <ticket-id> -- New investigation 1. **Check for existing investigation** -- if a `{ticket-id}` directory already exists under `.ai-engineering/support/`, resume the existing investigation rather than creating a duplicate. 2. **Create structure** -- create `.ai-engineering/support/{date}/{ticket-id}/` directory. 3. **Scaffold investigation** -- create `investigation.md` from template: ```markdown # {ticket-id}: {title} **Date**: YYYY-MM-DD **Customer**: {name/org if known} **Status**: investigating | resolved | escalated **Priority**: p1 | p2 | p3 ## Issue {Customer's description -- verbatim or summarized} ## Environment - Product version: - OS/Platform: - Configuration: ## S...