docs(infra): document Fuseki seed data init in stack README

Author: twicechild

src/infra/ted-sws-stack/README.md

@@ -67,6 +67,7 @@ docker compose -p ted-sws-stack logs -f
 | `.env.testing` | Testing/SRV passwords from Vault (generated by `make testing-unified-dotenv`) |
 | `airflow/Dockerfile` | Airflow image with celery provider |
 | `airflow/requirements.txt` | Python dependencies for Airflow |
+| `fuseki/seed/*.ttl` | Seed data loaded into Fuseki on first startup (one dataset per file) |
 
 ## Environment Configuration
 
@@ -165,6 +166,16 @@ docker volume rm minio-data mongodb-data fuseki-data sftp-data airflow-postgres-
 make start-local-stack
 ```
 
+## Fuseki Seed Data
+
+The `fuseki-init` service runs once on startup (after Fuseki is healthy) and
+idempotently creates datasets from TTL files in `fuseki/seed/`. Each file
+becomes a dataset named after the file (e.g., `test_limes.ttl` → dataset
+`test_limes`). Existing datasets are skipped.
+
+To add a new dataset, drop a `.ttl` file in `fuseki/seed/` and restart the
+stack. No code changes required.
+
 ## Migration from Old Setup
 
 ### Before (old way)