Starter templates for DriftQ — copy/paste production patterns for durable, replayable workflows.
These are intentionally small repos you can use as a baseline when you’re building real systems that need:
- retries + backoff
- DLQ (dead-letter queue)
- replay
- safe side effects (idempotency)
- versioned indexes + rollback (RAG)
FastAPI API + Next.js UI + worker, wired to DriftQ-Core (Docker Compose).
This is the “show it, don’t explain it” demo:
- force a failure
- watch retries
- DLQ payload persists
- replay with a fix applied → success ✅
- UI streams the run timeline live via SSE
Repo: https://github.com/driftq-org/driftq-starter-fastapi-nextjs Best for: quick demos, onboarding, proving the DLQ/replay story
A more production-ish starter that proves a real claim:
Retries won’t duplicate side effects — even if the worker crashes mid-step and DriftQ redelivers the message.
It shows a clean pattern:
- DriftQ for durable commands + redelivery
- idempotent worker
- exactly-once markers stored in SQLite (
effect_id) - simple proof artifacts you can eyeball
Repo: https://github.com/driftq-org/driftq-side-effects-starter Best for: payments/webhooks/tickets/LLM tool calls where double-execution is unacceptable
A small, production-minded demo that shows how to:
- use DriftQ-Core (via Docker) as a task queue + redelivery engine
- build and version a tiny Qdrant RAG index
- replay failed runs safely
- roll back an index by switching a Qdrant alias
Runs locally inside Docker (no local Python or jq needed).
Repo: https://github.com/driftq-org/driftq-rag-index-demo Best for: RAG ingestion/index builds, versioning + alias-based rollback, replayable pipelines
These starters are just templates. The engine lives here: