Testing · Version 1.0.0 · Reviewed 2026-08-02
Migration Parity Validator
Design confidence for dual-run parity design and behavioral equivalence testing with evidence, explicit trade-offs, and a verification plan.
4 method steps
6 documented failure modes
5 diagnostic checks
7 quality gates
Proves old and new implementations agree at the real integration boundary before a migration removes the fallback. It grounds the decision in paired outputs, routing, types, dimensions, timing, failure behavior, representative traffic, and tolerance rules and explicitly prevents unit tests passing while the new path drops, renames, reroutes, or aggregates data differently in production.
₹299 one-time
Get this skill archive
What it checks first
Migration Parity Validator proves old and new implementations agree at the real integration boundary before a migration removes the fallback. It grounds the decision in paired outputs, routing, types, dimensions, timing, failure behavior, representative traffic, and tolerance rules and explicitly prevents unit tests passing while the new path drops, renames, reroutes, or aggregates data differently in production. Use it when the work involves Dual-run parity design, Behavioral equivalence testing, Migration divergence analysis.
- Whether the old and new paths can coexist, which determines if incremental migration is possible at all.
- The true consumer inventory, including internal jobs, scripts, and integrations not visible in the main codebase.
- Data volume and the time the migration takes at production scale, not sample scale.
- Whether the change is backward compatible for data written by the previous version during rollout.
- The rollback path, and specifically whether it remains available after the first irreversible step.
Example task
Input
Apply the migration parity validator to our current dual-run parity design work. We need a concrete decision, bounded changes, and evidence that the result is correct.
Expected output
Start with paired outputs, routing, types, dimensions, timing, failure behavior, representative traffic, and tolerance rules. The highest-risk failure is unit tests passing while the new path drops, renames, reroutes, or aggregates data differently in production. Compare both paths on the same events at the boundary consumers observe and classify every divergence. Verify the result by running a representative replay with forced old-path, new-path, and dual-path failures.