What problem does it solve? Migrating a dbt project from dbt-core to dbt v2 produces dozens of unfamiliar errors, and users cannot tell which ones they can fix themselves versus which are blocked on v2 engine updates. This Skill runs dbt-autofix first, then classifies every remaining error into four actionable categories so migration work proceeds in the right order. ## Core Features & Use Cases - Mandatory triage procedure: Enforces a strict execution order — optional dbt debug credential check, dbt-autofix run with git diff review, then error classification — before any manual fixes are proposed. - Four-category classification framework: Sorts errors into auto-fixable (quote nesting, static analysis in analyses/), guided fixes with diffs (config API changes, source name mismatches, case-sensitive identifiers), needs-user-input (hardcoded FQNs, failing analyses), and blocked-on-v2 (engine gaps, MiniJinja differences, adapter methods not implemented). - GitHub issue lookup: Proactively searches dbt-labs/dbt-fusion issues via the GitHub API for suspected v2 engine bugs and links them in the report. - Use Case: A data engineer upgrades to dbt v2 and dbt compile --static-analysis strict fails with 15 errors. The Skill runs autofix, reviews its changes, then reports 4 auto-fixable issues, 6 guided fixes with diffs, 2 needing decisions, and 3 blocked on tracked v2 issues. ## Quick Start Help me triage the errors from migrating my dbt project to dbt v2 by running dbt-autofix and classifying what remains.
