Features

Map it. Understand it. Run it.

Merethold brings topology, health, impact, procedures and the business context around your infrastructure into one fast, keyboard-friendly workspace.

See

A living, editable topology

One durable model of assets and relationships drives the map, the inventory, monitoring and search.

Full-screen topology map

Regions, networks and subnets group automatically. Zoom from the whole estate to a single container, with level-of-detail cards that stay fast past 500 assets.
In-place asset detail

Open any server without leaving the map: network interfaces, addresses, VLANs, firewall rules, services, backups, changes and history.
Typed relationships

routes to, runs on, depends on, replicates to, backs up to, protects, connects to: each with provenance, so you know whether a link was discovered, imported or drawn by hand.
Measured link traffic

Throughput per link from real interface counters, so the busiest paths stand out at a glance.
Keyboard-first

Search by hostname or IP, arrow between assets, trace and impact with a key press, undo and redo edits. Ctrl K from anywhere.
Safe to edit

Saved views store layout only. Manual notes and positions survive every re-sync, and missing sources are flagged as stale for review rather than deleted.

Understand

From any component to the applications, products and customers that depend on it, with the path that explains why.

What breaks if this goes down

Select an asset and see applications down or at risk, the products they serve and the customers each product affects.
Confirmed failover only

Redundant paths only count when failover has been confirmed with evidence, such as a drill or a change record. Hope is not a strategy, and the map says so.
Incident snapshots

When an incident opens, Merethold records the impact at that moment, so post-incident reviews start from facts.
Customer-aware

Customers ↔ products ↔ infrastructure. Account staff can see who is affected without seeing network internals.

Operate

Health, procedures, history and money in the same place as the map.

Live health and telemetry

Up/down, CPU, memory, disk, network and uptime via Grafana and Prometheus, plus firing alerts. Every value records its source and timestamp.
Runbooks and procedures

Attach procedures to assets, open them in context and record each run. Back to the map returns you exactly where you were.
Incidents, changes and audit

Track incidents and changes against the assets they touch, with a full audit log of who did what.
Renewals and costs

Domains, certificates, servers and subscriptions with expiry dates, suppliers and what each one pays for, synced from your providers and accounting.
Imports without duplicates

CSV and JSON import with a dry-run preview. Re-importing the same file changes nothing.
Read-only integrations

Monitoring, cloud, DNS and Zero Trust, containers, proxies, storage, code, errors, uptime, on-call, tickets, billing and accounting.

Automate safely

Give Claude and other MCP clients your infrastructure context, and keep humans in charge of access.

MCP server

Query assets, topology, health and live telemetry, impact, alerts, incidents, changes, renewals, backups and runbooks over Streamable HTTP with personal keys.
Short-lived SSH certificates

Ed25519 certificates for one server and one login, 15 minutes by default, hard-capped at an hour, bound to the requester’s address.
Look, Fix and Admin tiers

Read-only access can be automatic under standing rules; Fix needs an open incident assigned to the requester; Admin always needs an owner’s approval.
Accountable by default

Daily rate limits pause runaway agents, and every login is traceable to the request, reason and approver.

Already have an account? Sign in