Driftn Documentation

Audits Vercel, Supabase and/or AWS for cost waste and optimisation opportunities. At least one API key is required.

# Example: Vercel + Supabase audit
curl -X POST .../audit \
  -H "X-API-Secret: YOUR_SECRET" \
  -d '{"vercel_key":"vk_live_...", "supabase_key":"eyJ..."}'

Runs the same audit as POST /audit and returns a branded PDF document instead of JSON. Accepts the same parameters.

Runs an audit then generates a 30/90-day cost forecast based on historical data. Accepts the same body as /audit.

# Response includes both audit + prediction
{
  "audit": { ... },
  "prediction": {
    "next_30_days": 310.00,
    "next_90_days": 980.00,
    "trend": "increasing",
    "confidence": 0.82
  }
}

Calculates a 0–100 Driftn Score based on cost efficiency, deployment health, security posture and prediction risk. Stored in Supabase if user_email is provided.

Applies a suggested remediation from an audit or score result — for example, removing an idle Lambda function or fixing a missing security header. The set of allowed actions is restricted server-side via an allowlist.

Runs a security scan using the vErtex engine. Covers 12 modules including SSL, OWASP Top 10, headers, WAF, CMS detection, exposed files and API endpoints.

Check SSL certificates for up to 20 domains at once.

{ "domains": ["example.com", "api.example.com"] }

Analyses HTTP security headers and generates a remediation config. Pass generate: "vercel" or "nginx" to get a ready-to-paste config file.

{
  "url": "https://example.com",
  "generate": "vercel"
}

Returns recent deployments from a connected Vercel project, including status, duration, and commit metadata. Used to power the Pipeline Health and overview deploy list in the dashboard.

Pings a URL from the backend server. More reliable than browser-based pings — avoids CORS and client-side network issues.

Returns all endpoints registered for monitoring for a given user_id.

GET /api-monitor/endpoints?user_id=default

Returns the last limit ping results (default 50) for an endpoint, ordered newest first.

Returns all nodes and edges for the authenticated user. Use /infra/graph/demo (no auth) to preview the graph shape with sample data.

Re-derives nodes and edges from your connected Vercel/Supabase/AWS credentials and emits a graph_sync event onto the timeline.

Supports filtering by severity (critical/high/medium/low/info) and type (deployment, cpu_spike, api_error, incident_opened, etc). /infra/events/demo returns a realistic 10-event incident chain with no auth required — useful for previewing the Root Cause Timeline UI.

Recursively follows an event's related_ids (max depth 10) and returns the full chain sorted chronologically — e.g. deploy → CPU spike → memory spike → Redis timeout → API errors → rollback → recovered.

Returns lightweight snapshot summaries (not the full payload) ordered newest first. Use POST /infra/snapshots to create one on demand.

Compares two snapshots and returns added/removed/modified nodes and edges, plus the cost and health deltas between them. Powers the "Yesterday → Today" comparison view.

Returns the Driftn Score and savings history over the last days (default 30, max 365), plus a count of critical events in that window. Powers the trend charts in the dashboard overview.

Iterates over users with monitoring_enabled = true, pings their tracked endpoints, and emits api_error or latency_spike events for failures. Recommended schedule: every 6 hours.

Sends each monitored user an email with their week's event count, critical alerts, deploys, and incidents. Recommended schedule: Mondays at 9am.

Toggles continuous monitoring, email alerts, and the weekly digest for a user. Emits a config_change event so the change shows up in the activity feed.

Returns tasks filtered by user_id (default) or team_id. Also available: POST, PATCH /{id}, DELETE /{id}.

Returns events ordered by date. Event types: task | deadline | meeting | release.

Returns up to limit messages (default 50) for a channel and team_id. Use POST /teamspace/messages to send.

Returns live MRR, ARR, churn rate and per-plan breakdown from your Stripe account. If STRIPE_SECRET_KEY is not configured, returns mock data with "source": "mock".

Returns active Stripe subscriptions with plan, amount, and start date — the customer-level breakdown behind the /mrr summary.

Receives Stripe events and logs them to Supabase. Configure in your Stripe Dashboard → Webhooks pointing to this endpoint. Requires STRIPE_WEBHOOK_SECRET.

Handled events: customer.subscription.created, customer.subscription.updated, customer.subscription.deleted, invoice.payment_failed.

Adds an email to the beta waitlist and sends a confirmation email via Resend. Idempotent — calling it again with the same email returns {"status": "already_registered"} instead of an error.

{ "email": "dev@example.com" }

Creates a Stripe Checkout session in one-time payment mode (not a subscription). Limited to 50 seats — once claimed, the endpoint returns 409.