# AdaptAI Data Transfer Operations (DTO) Framework Manifest‑driven, event‑based operations platform for data migrations and transfer workflows. Includes generators, CI validation, service orchestration, integrations, and operational runbooks. ## Quick Start - Validate configuration: `python validate-ci.py` - Generate artifacts: `python generate.py --environment production` - Manage services: `python services/dto_service_manager.py --help` ## Core Concepts - Manifest: `dto_manifest.yaml` defines services, SLOs, ports, overlays, and policies. - Generation: `generate.py` renders supervisord configs and per‑service runbooks from the manifest. - Validation: `validate-ci.py` enforces schema, port conflicts, deps, SLOs, CODEOWNERS, overlays, and generated artifacts. - Orchestration: Supervisord manages DTO services; health/monitor utilities live under `services/`. - Events: NATS JetStream backbone with topics and consumers under `events/`. - Storage: Dragonfly (cache), ScyllaDB + JanusGraph (lineage/durable), optional Redis cluster spec in `database/`. - Integrations: Slack/Jira/Confluence and Hugging Face Hub (Xet backend) under `integrations/`. ## Common Commands - Validate: `python validate-ci.py` - Generate all: `python generate.py --environment production` - Service manager: `python services/dto_service_manager.py status` - Make targets: `make validate`, `make generate`, `make run`, `make security` ## Environments & Overlays Environment overlays live in `overlays/{development,staging,production}` and are referenced by the manifest’s `environments` section. Use `--environment` to target specific overlays during generation. ## Secrets & Security - Do not commit secrets. Keep `.env` untracked (see `.gitignore`). - Required env vars for HF integration include `HF_TOKEN`, `HF_REPO_*`. See `docs/xet_lfs_user_guide.md` and `integrations/huggingface_config.yaml`. - Security docs: `security_alert_summary.md`, `security_response_plan.md`. ## Directory Map - `dto_manifest.yaml` — Service catalog and platform policies - `generate.py`, `templates/` — Supervisord/runbook generators - `validate-ci.py`, `schema/` — CI validation gates & schemas - `services/` — Supervisord config, health/crash/memory monitors, manager - `events/` — NATS config, event schema, producers/tests - `integrations/` — Slack/Jira/Confluence, Hugging Face (Xet) - `cache/`, `database/`, `lineage/` — infra configs and clients - `overlays/` — environment overlays - `docs/` — runbooks and architecture docs - `scripts/` — archive protocol and migration utilities ## Notes on Supervisord The generated supervisord template maps DTO services to the actual script locations: - `dto-lineage-handler` → `lineage/lineage_event_handler.py` - `dto-slack-automation` → `integrations/slack_bot.py` - `dto-jira-automation` → `integrations/jira_automation.py` - `dto-confluence-automation` → `integrations/confluence_automation.py` - `dto-jira-webhooks` → `integrations/jira_webhooks.py` - `dto-health-monitor` → `services/health_monitor.py` Dragonfly cluster is managed externally and not emitted as a supervisord program by default. ## Testing Minimal tests are included under `tests/`: - HuggingFace client path‑sanitization - Manifest validation via generator Run: `pytest -q`