add parity test on onboarding

Signed-off-by: Alexandros Koumparoulis <akoumparouli@nvidia.com>

Author: akoumpa

.claude/skills/model-onboarding/SKILL.md

@@ -298,6 +298,23 @@ Add fully commented example configs under `examples/`:
 
 ---
 
+## Phase 6: Parity Testing
+
+After implementation and unit tests are complete, run the full parity-testing
+workflow to verify that the new model produces numerically equivalent results to
+the reference HuggingFace implementation.
+
+**Read and follow the parity-testing skill** at
+`.claude/skills/parity-testing/SKILL.md`. It walks through three levels of
+comparison (state-dict round-trip, component-level parity, end-to-end forward
+pass) and provides debugging steps when a level fails.
+
+Do not skip this phase. A model that passes unit tests can still diverge from HF
+due to subtle weight-conversion bugs, backend differences, or RoPE mismatches
+that only surface in a full parity comparison.
+
+---
+
 ## Key Files Reference
 
 | File | Purpose |
@@ -336,4 +353,5 @@ Add fully commented example configs under `examples/`:
 - [ ] Created layer equivalence tests for every rewritten layer (matching model dtype)
 - [ ] Created functional tests (training loss decreases)
 - [ ] Updated docs/model-coverage page
+- [ ] Ran parity-testing skill (state-dict round-trip, component parity, E2E forward pass)
 - [ ] Set `ModelClass = <Name>ForCausalLM` at module bottom