# Clinical Trials 1.0.0 — quickstart

Synthetic study, subject, visit, safety, randomisation and central-lab test data. This pack cannot be activated yet (NOT_IN_ANY_PLAN, NOT_IN_AGENT_IMAGE, NOT_IN_CONTROL_PLANE_CATALOGUE, NO_HOSTED_SANDBOX).

All data in this bundle is synthetic, generated from a fixed seed by the pack's own generator. Real data is processed only by the DataNivra agent inside your network (customer-resident processing); this bundle never contains customer data, keys or credentials.

## Prerequisites

- This pack is not in any plan yet.
- One DataNivra agent inside your network (outbound HTTPS only) that ships Clinical Trials 1.0.0.
- Read-only access to a compatible source (verified with this pack: PostgreSQL, Local files (Parquet / CSV)).
- A non-production target environment the agent may write test data to.
- A masking key in your own secret store, referenced as vault://…, azure-kv://… or env://… (DataNivra only ever sees the reference).
- For production sources, a second person who approves policies (separation of duties).

## First certified dataset

1. **Try it with synthetic data.** Sign up and start the hosted sandbox (`POST /v1/sandbox`,
   `api/sandbox.json`, or `datanivra sandbox start`). Nothing to install.
2. **Inspect and activate the pack.** `GET /v1/industry-packs/clinical-trials` shows versions,
   entitlement, activation checks and which of your agents can run it;
   `POST /v1/industry-packs/clinical-trials/enable` (or `datanivra packs enable clinical-trials`) activates it.
3. **Install or reuse the agent.** One agent runs many packs. It must report
   `clinical-trials` 1.0.0 with the catalogued integrity digest
   (see `docs/runbooks/deploy-agent-docker.md`).
4. **Connect a source** through the normal source workflow (connection *reference* only).
   Verified with this pack: PostgreSQL, Local files (Parquet / CSV).
5. **Choose scenarios**, for example:
   - `CT_ROUTINE_CONDUCT` (NORMAL): Ordinary study conduct: enrolled subjects, in-window visits, forms, labs, randomisation and kits
   - `CT_SCREEN_FAILURES` (RARE): Subjects who fail eligibility at screening, with the criterion that was not met
   - `CT_SERIOUS_ADVERSE_EVENTS` (RARE): Serious, severe adverse events with safety case reports, some unexpected
6. **Create policy drafts from the pack templates:** `POST /v1/industry-packs/clinical-trials/policy-drafts`
   (`api/policy-drafts.json`). Review, submit and approve them (a second person approves policies
   for production sources).
7. **Request the dataset** (`api/dataset-request.json`). The agent subsets, masks, generates the
   scenarios, validates and **certifies** it inside your network.
8. **Provision** the certified version to a non-production environment (`api/provision.json`).
9. **Download the evidence**: certification report and evidence references
   (`GET /v1/certification-reports?dataset_id=…`, `GET /v1/evidence?dataset_id=…`).

See `TROUBLESHOOTING.md` for reason codes and recovery.