Validic Inform connects data from hundreds of health apps, wearables, and in-home medical devices, and delivers it to you in one normalized data model. This repo is the fastest way to see that working — two small apps that exercise the API end to end, so you can watch real calls succeed before you write a line of your own integration.
| App | Best for | Stack | Setup |
|---|---|---|---|
| quickstart | Seeing the four core API calls work in about 5 minutes | Node.js + Express, vanilla HTML | Org ID + token, nothing else |
| explorer | A full API workbench — build requests, inspect responses, generate synthetic test data | Next.js + React | Org ID + token; Data Generator URL optional |
New to Inform? Start with quickstart to learn the shape of the API. Move to explorer when you need to poke at every endpoint or test your data handling against realistic payloads.
Local development tools only. These sample apps are meant to run on your machine. The explorer app's API routes proxy requests using your credentials, so don't deploy it to a public URL as-is.
- Node.js 22+
- An Inform account — sign up free at dashboard.validic.com
- Your Org ID and Org Token, both visible on your dashboard
1. Clone the repo
git clone https://github.com/validic/inform-quickstart.git
cd inform-quickstart2. Add your credentials
For the quickstart app:
cp apps/quickstart/.env.example apps/quickstart/.envOpen apps/quickstart/.env and fill in:
ORG_ID=your_org_id_here
ORG_TOKEN=your_org_token_here
For the explorer app:
cp apps/explorer/.env.example apps/explorer/.envOpen apps/explorer/.env and fill in:
ORG_ID=your_org_id_here
ORG_TOKEN=your_org_token_here
API_URL=<contact your Inform representative>
DATAGEN_URL=<contact your Inform representative>
STREAM_URL=<contact your Inform representative>
DATAGEN_URL is optional — without it, everything works except the explorer's Generate Data panel. Your .env files are gitignored; keep them that way. The API authenticates with a token query parameter, so treat your token like a password — it will appear in URLs, logs, and any curl command you copy.
3. Run an app
npm run quickstart # installs dependencies and starts the quickstart app# or
npm run explorer # installs dependencies and starts the explorer appOn boot, the quickstart app provisions what it needs and tells you so:
Checking for existing Inform users...
Using existing user: 59b02bdef758800001e13e8c
Checking for existing Inform streams...
Stream created: 6a4ec486baefa8000148a676
Inform Sample App running at http://localhost:3000
4. Open http://localhost:3000
Four panels, four core API capabilities. Every panel has a copy curl button that gives you the exact request the app just made — run it yourself, change it, break it, learn from it.
| Panel | API call | What it demonstrates |
|---|---|---|
| Connect Your Device | GET /organizations/:org/users/:uid | Connected sources, read from the user's sources array; links to the device marketplace |
| Your Data | GET .../users/:uid/:type | 30 days of normalized data — summaries, workouts, sleep, measurements |
| Live Stream | GET /streams/:id/connect (SSE) | Events arriving in real time |
| Replay last 30 days | GET /replay (SSE) | Historical events through the same stream interface |
Together they walk the pattern every real integration follows:
- Provision a user — you get back a
marketplace.urlfor that user - Connect a device — open the marketplace URL and authorize a source (the app does this via the "Connect a device or app" button)
- Wait a moment — first sync can take a couple of minutes depending on the source; an empty
dataarray right after connecting is normal - Verify two ways — pull the data over REST and confirm the same event arrived on your stream
The full testing walkthrough — including device-specific steps and generating synthetic test data — lives in the guides at developer.validic.com.
A three-panel workbench: API setup on the left, request builder in the middle, live response viewer on the right.
- Every users endpoint — get, list (with pagination and sorting), provision
- Every data type — summaries, workouts, sleep, nutrition, measurements, intraday, CGM
- Stream replay over SSE
- Generate Data — create realistic synthetic records for a user without owning the device (requires a Data Generator URL from your Inform representative)
- Live request preview (method, URL, headers, body) before you execute, and copy-as-curl on everything
inform-quickstart/
├── apps/
│ ├── quickstart/ # Express app — four core calls, minimal code to read
│ │ ├── server.js # every API call the app makes, in one file
│ │ ├── public/ # vanilla HTML/JS frontend
│ │ └── tests/
│ └── explorer/ # Next.js API workbench
│ └── src/
├── agents.txt # pointers for AI coding agents
└── package.json # npm run quickstart / npm run explorer
| Symptom | What's going on |
|---|---|
404 from GET /users/<id> | The single-user endpoint is keyed by uid (the identifier you chose at provisioning), not the Validic id. Data endpoints accept either. |
422 maximum number of streams exceeded on boot | Orgs are capped at 5 streams. The app still starts (live streaming disabled) and the logs print the curl commands to list and delete old streams. |
Empty data right after connecting a device | Expected — first sync can take up to a couple of minutes depending on the source. |
| Explorer asks for a config upload | Its .env needs ORG_ID, ORG_TOKEN, and API_URL at minimum. DATAGEN_URL is optional and only gates the Generate Data panel. |
| Lots of requests failing at once | Rate limit is 2,000 requests per token per hour — fine for manual testing, worth knowing if you script anything. |
- API reference and guides:developer.validic.com
- For AI agents:developer.validic.com/llms.txt — every docs page in Markdown plus the OpenAPI spec
- Dashboard / free sandbox:dashboard.validic.com
MIT — see LICENSE.