Getting Started

Start with the core service

Arcnem Vision ships a Flutter demo client, but the fastest way to understand the product is through the dashboard and upload API. The core loop is: ingest an image, run a workflow, inspect the results, and review the run trace.

Prerequisites

• Docker + Docker Compose
• Bun
• Go 1.25+ (agents, MCP)
• CompileDaemon (go install github.com/githubnemo/CompileDaemon@latest)
• Flutter SDK
• Tilt

tilt up launches the Flutter demo client too, so the Flutter SDK is still part of the default local stack. If you are evaluating the platform, focus on the dashboard and server-side services first.

1. Clone and configure

git clone https://github.com/arcnem-ai/arcnem-vision.git
cd arcnem-vision

Copy every .env.example to .env:

cp server/packages/api/.env.example server/packages/api/.env
cp server/packages/db/.env.example  server/packages/db/.env
cp server/packages/dashboard/.env.example server/packages/dashboard/.env
cp models/agents/.env.example       models/agents/.env
cp models/mcp/.env.example          models/mcp/.env
cp client/.env.example              client/.env

Add your provider keys:

• OpenAI API key → OPENAI_API_KEY in models/agents/.env
• Same OpenAI key (recommended) → OPENAI_API_KEY in server/packages/api/.env for dashboard collection chat and AI workflow draft generation
• Replicate API token → REPLICATE_API_TOKEN in models/mcp/.env

Everything else is already configured for local development. Postgres, Redis, and MinIO come from docker-compose.yaml.

2. Start everything

tilt up

Tilt installs dependencies, starts infrastructure, runs migrations, and launches the API, dashboard, agents, MCP server, Inngest, docs site, and the Flutter demo client. Open the Tilt UI at http://localhost:10350 for logs and manual resources like seed and introspection.

3. Seed the database

In the Tilt UI, trigger seed-database.

The seed creates:

• a demo organization, project, workflow keys, service keys, and API keys
• editable workflows and reusable workflow templates
• sample images for the description, OCR, quality-review, and segmentation paths
• stored OCR results, descriptions, embeddings, segmentations, and example run history
• a local debug dashboard session

Because server/packages/api/.env.example enables API_DEBUG=true, the dashboard can bootstrap into the seeded local session after seeding.

4. Walk the core product

1. Open the dashboard at http://localhost:3001.
2. In Projects & API Keys, inspect seeded workflow keys, service keys, and their attached workflows.
3. In Workflow Library, browse templates, click Generate With AI, or open a graph in the canvas.
4. In Docs, inspect seeded documents or upload a new one from the dashboard.
5. In Runs, open a run and inspect its initial state, per-step deltas, final state, timing, and errors.

5. Exercise the two ingestion paths

Workflow / API-key path

Use a workflow API key to run the automated flow:

curl -X POST http://localhost:3000/api/uploads/presign \
  -H "Content-Type: application/json" \
  -H "x-api-key: ${API_KEY}" \
  -d '{"contentType":"image/png","size":12345}'

Then upload to the returned S3 URL and call /api/uploads/ack. That acknowledgement verifies the object, creates the document, and queues document/process.upload for the workflow key’s bound workflow.

Dashboard path

In the Docs tab:

1. Click Add From Dashboard.
2. Upload an image into a project.
3. Open the saved document.
4. Queue any saved workflow against it.

This path is useful for ad-hoc analysis, reruns, and operator-driven evaluation because the document is not tied to a workflow key by default.

Health checks

GET http://localhost:3000/health   # API
GET http://localhost:3020/health   # Agents
GET http://localhost:3021/health   # MCP

S3 config details

Default local dev uses MinIO from docker-compose.yaml. The .env.example files ship with working defaults:

• S3_ACCESS_KEY_ID=minioadmin
• S3_SECRET_ACCESS_KEY=minioadmin
• S3_BUCKET=arcnem-vision
• S3_ENDPOINT=http://localhost:9000
• S3_REGION=us-east-1
• S3_USE_PATH_STYLE=true

For hosted storage, substitute your AWS S3, Cloudflare R2, Railway Object Storage, or Backblaze B2 credentials.

• Set S3_USE_PATH_STYLE explicitly for your provider.
• Cloudflare R2 commonly needs S3_REGION=auto and S3_USE_PATH_STYLE=false.
• When the dashboard uploads directly from the browser to storage, some providers such as R2 also need bucket CORS configured to allow your dashboard origin and PUT requests.