from dlthub-ai-workbench41
Diagnose and fix dlthub transformation failures: SQL dialect incompatibility, pipeline state errors, and missing columns after deployment.
Provides step-by-step diagnostics and remediation for dlthub transformation failures. It helps detect SQL dialect portability issues, recover pipelines stuck in failed states, and identify/repair missing or silently dropped columns caused by type inference. The skill documents static compatibility checks, escalation steps for pipeline recovery, and destination-specific workarounds (e.g., DuckDB).
Use this skill when a transformation works locally but fails on another destination (BigQuery, Snowflake, Postgres), when pipelines exhibit syntax/unsupported-function errors, when runs are stuck in retry loops, or when expected columns are absent from the output. Also use proactively to run static SQL compatibility checks before deployment.
Best used by agents with shell/Bash and dlt CLI access; compatible with tooling that can run static SQL checks (SQLGlot) and invoke pipeline commands.
This skill has not been reviewed by our automated audit pipeline yet.
