chore(community): add good first issue template for normalization rows

.github/ISSUE_TEMPLATE/normalization_row_task.md

@@ -0,0 +1,38 @@
+---
+name: Normalization row task (good first issue)
+about: Add new Pashto text-normalization rows to the seed dataset
+title: "[Data][Normalization] Add rows for "
+labels: data, good first issue
+assignees: ''
+---
+
+## Goal
+Add new high-quality rows to:
+- `data/processed/normalization_seed_v0.1.tsv`
+
+## What to add
+- [ ] 5-20 new rows with unique `id` values
+- [ ] `raw_text` and `normalized_text` filled for every row
+- [ ] short `note` explaining the normalization change
+
+## Quality checklist
+- [ ] no empty fields
+- [ ] no duplicate `id`
+- [ ] punctuation/spacing normalized consistently
+- [ ] meaning preserved (no semantic rewrite)
+
+## Validation
+Run:
+```bash
+python scripts/validate_normalization.py data/processed/normalization_seed_v0.1.tsv
+```
+
+Paste result here:
+```
+# validation output
+```
+
+## Acceptance criteria
+- [ ] validator passes
+- [ ] rows are realistic Pashto examples
+- [ ] PR links this issue