Service catalog & uptime monitoring for engineering teams
An API-first service catalog with built-in uptime monitoring. Register a service, run a check, route an incident — all from one API. Your agents can drive it too.
API / CLI / SDK / MCP
Services, blueprints, checks — every entity is a row you write to via a REST API or one of four clients. Run it from CI, your terraform, your shell, or your agent. The platform monitors what you registered.
# Register a service. The API key sets your workspace.
curl -X POST https://api.wanepia.com/v1/entities \
-H "Authorization: Bearer wnp_live_•••••••••••••" \
-H "Content-Type: application/json" \
-d '{
"blueprint": "service",
"slug": "payments-api",
"team": "payments",
"checks": [
{ "kind": "http", "url": "https://api.acme.io/health",
"expect": 200, "every": "30s" }
],
"notify": ["#payments-oncall", "pager:payments"]
}'
# 201 Created · monitoring starts immediately.# One binary. Reads ~/.wanepia/config or $WANEEPIA_TOKEN.
$ wnp entity new payments-api \
--blueprint service --team payments \
--check http://api.acme.io/health --every 30s
✓ entity payments-api created (srv_a2qPx)
✓ check http://api.acme.io/health · every 30s
✓ notify #payments-oncall, pager:payments
$ wnp status --team payments --watch
payments-api ● up 99.99% 34 ms
checkout-web ● degraded 98.21% 812 ms
invoicing ● down — timeout# claude_desktop_config.json · drop-in MCP server.
{
"mcpServers": {
"wanepia": {
"command": "npx",
"args": ["-y", "@wanepia/mcp"],
"env": { "WNP_TOKEN": "wnp_live_•••" }
}
}
}
# Your agent gets these tools:
list_entities // search the catalog
get_entity // full record + recent checks
open_incident // flip state, page on-call
acknowledge // claim an active incident
query_status // uptime % by team, windowPLATFORM
Most teams stitch this together from four tools. Catalog in a wiki, checks in one SaaS, on-call in another, status page in a fourth. We collapse the seams.
Define the shape of a service, a job, a datastore — any noun your team owns. Each blueprint gets its own properties, defaults, and required checks. Inheritance, not free text.
HTTP, TCP, DNS, TLS, and agent push for services on private networks. Run every 10s up to every hour, with body-match and failure thresholds. Services behind firewalls, VPNs, or inside Kubernetes push results in — no exposure needed.
Every status change — up → degraded → down → up — is recorded as an immutable event. Multi-check consensus prevents false positives when one check out of many fails.
Slack, Discord, webhook, NATS subjects, and ntfy.sh push notifications. Per-check cooldown, recovery notifications, and repeat alerts while an entity stays down. Silence without deleting the policy.
Create named API keys for CI, the CLI, or your agents. Invite teammates by email — admins manage keys and members; every session is attributed to a specific user.
Your catalog is the source of truth your agents need. Drop the MCP server into Claude, Cursor, or your own runtime and the same scoped tokens that gate humans gate the model.
One click to publish a public status page for your stakeholders. Hash-based URL keeps your tenant identity private. Real-time entity grid, incident timeline, and configured alert channels — all read-only.
Model relationships between services with depends_on,
calls, and parent_of edges.
Trace blast radius from a single failing entity across your entire catalog
before an incident widens.
HEALTH / LAST 24H
One row per entity, one cell per hour. Recurring flaps, time-of-day correlations, the noisy neighbour — they all surface the moment you stop looking at a hundred line charts.
ROADMAP
The core catalog, checks, heatmap, MCP server, and public status pages are live. Here's what we're building next — get notified when something ships.
Want to shape the order? Tell us what you need.
WHY WANEPIA
Most teams outgrow monitoring tools before they're ready for a full IDP. Wanepia lives in that gap — a catalog-first platform that adds monitoring, not the other way around.
Great monitoring tools — but monitoring-first. There's no service catalog, no entity relationships, no API to drive from CI or your agent. You end up with dashboards that don't map to your architecture.
Powerful, but it's a full internal developer portal. If all you need is a service catalog you can query and monitor, Backstage means weeks of setup, a plugin ecosystem to maintain, and a Kubernetes cluster to run it on.
Status pages and incident comms done well — but they're the output layer, not the source of truth. No catalog, no entity graph, no API you can write to from your deploy pipeline or agent.
PRICING
Bring as many teammates as you like — humans and agents both. You're charged for the things in the catalog, not the people looking at them.
Hobby projects, side services, a homelab you'd like to keep alive.
Working teams that need more capacity, public status pages, and richer integrations.
SSO, audit logs, single-tenant region, contracts your legal team will love.
wanepia — always running, always with you