# mountOS

> Self-hosted, POSIX-compatible distributed storage with native macOS, Linux, and Windows mounts plus S3 and HDFS, and built-in forks, time travel, and one-minute versioning.

This file helps AI agents discover, understand, deploy, and get support for mountOS. Each link below points to a single self-contained topic. The concatenated corpus is at https://mountos.io/ai/llms-full.txt.

## Docs

- [Overview](https://mountos.io/ai/topics/overview.md): Self-hosted, POSIX-compatible distributed filesystem with native macOS/Linux/Windows clients plus S3 and HDFS, built-in forks, time travel, one-minute versioning, per-volume retention, and a read-only MCP connector for AI agents
- [Architecture](https://mountos.io/ai/topics/architecture.md): How mountOS splits into metadata/client/gateway planes, regions and clusters, dataserv ownership, client backends, and the custom encrypted transport
- [Components](https://mountos.io/ai/topics/components.md): Catalog of mountOS deployable server services, per-OS mount clients, SDKs, and admin tooling a third party integrates with
- [Deployment](https://mountos.io/ai/topics/deployment.md): How to deploy a self-hosted mountOS suite from official binaries (download.mountos.sh). Covers the service fleet, external dependencies, the network and HA model, the bring-up order, and the shipped client-side assets (CSI, systemd)
- [Configuration and Environment](https://mountos.io/ai/topics/configuration.md): Configuration surface for mountOS server components, build-time identity, and the mountos client flags, grouped by component with verified defaults
- [Database and Schema](https://mountos.io/ai/topics/database-schema.md): How mountOS stores control-plane and metadata state in SQL, the supported engines and capability routing, schema split, repo layout, and the schema generation and migration model
- [Vaults and Secrets](https://mountos.io/ai/topics/secrets.md): How mountOS services read operator-supplied secrets from a pluggable vault, with cross-backend version identity, key kinds, topology, permissions, and rotation
- [Download and Install](https://mountos.io/ai/topics/download-install.md): Public download and install guide covering installer scripts, packages, artifact layout, and source builds
- [Support and Compatibility](https://mountos.io/ai/topics/support.md): How to obtain mountOS via the public installer, its versioning and release matrix, and protocol/S3/WebHDFS compatibility expectations
- [Getting Started (Build a System)](https://mountos.io/ai/topics/getting-started.md): End-to-end setup from nothing to a mounted volume, standing up the HUB, provisioning with the admin SDK, creating a region and its default cluster, deploying the in-region services, and creating volumes and access keys
- [Admin SDK and API](https://mountos.io/ai/topics/admin-sdk.md): The open connector to a running deployment's HUB, Ed25519-signed Admin API access through TypeScript, Go, and Rust packages or REST, with the response envelope, pagination, error model, and client generation from api.yaml
- [Change Events](https://mountos.io/ai/topics/change-events.md): Per-volume ordered change-event feed (create, delete, modify, rename) for general volumes, enabled by eventLogRetentionPeriod, read with mountos event as a stdout stream or a local HMAC-authenticated REST API, with capture, pagination, and consumer semantics
- [Troubleshooting](https://mountos.io/ai/topics/troubleshooting.md): Diagnose with the built-in tooling, the mountos check command (per-platform backend verification with printed fixes), dashboards, the HUB alerts view, and a symptom-to-fix log reference for registration refusals, boot exits (env, license, vault), db install/migrate, client mount and discovery, gateway rejections, and change-event auth

## Resources

- [Agent skill](https://mountos.io/skill.md): drop-in instructions for an AI agent (read this, then act)
- [Download](https://download.mountos.sh): release artifacts and install instructions
- [Support](https://mountos.io/support): how to get help
- [Full corpus](https://mountos.io/ai/llms-full.txt): every topic concatenated