Quickstart

Try it in two minutes

git clone https://github.com/NoplagLabs/noplag-engine
cd noplag-engine
docker compose up

First start migrates the schema and indexes the bundled sample corpus (~1,000 Wikipedia articles — takes a couple of minutes). Then:

• http://localhost:8000 — demo page. Paste a paragraph from a well-known Wikipedia article (say, the lead of Albert Einstein) and run a check: you get a similarity score, highlighted passages, and the matched articles.
• http://localhost:8000/docs — interactive OpenAPI docs for the /v1 API.

The demo page is deliberately bare — the API is the product here. Build your own UI on POST /v1/checks → GET /v1/checks/{id}/report.

API in 30 seconds

# submit
curl -s -X POST http://localhost:8000/v1/checks \
  -H 'Content-Type: application/json' \
  -d '{"query_text": "Text to check, at least a few sentences long..."}'
# -> {"check_id": "...", "status_url": ..., "report_url": ...}
 
# poll status, then fetch the report
curl -s http://localhost:8000/v1/checks/<id>
curl -s http://localhost:8000/v1/checks/<id>/report

The full surface (uploads, SSE progress, corpus management) is covered in the API reference and served live at /docs on your instance.

A note on authentication

The API ships with no authentication — it is a single-tenant backend service. Run it on a trusted network or put your own auth in front. See the security model before deploying anywhere reachable.

Next steps

• Self-hosting guide — compose and bare installs, operations, scaling.
• Bring your own corpus — the engine only matches against what you index.
• Configuration — every environment variable, annotated.