# RubixKube Docs

> Site Reliability Intelligence for modern infrastructure. Detect anomalies, diagnose root cause, and resolve failures across Kubernetes, AWS, GCP, and Linux VMs.

- [Welcome to RubixKube](https://docs.rubixkube.ai/getting-started/overview.md): RubixKube is the Reliability Layer for the AI era. An AI-native mesh of agents that watches your infrastructure, diagnoses root cause, and keeps systems alive across Kubernetes, AWS, GCP, and Linux VMs.
- [Quickstart](https://docs.rubixkube.ai/getting-started/quickstart.md): The shortest path to RubixKube. Create an account, connect an environment, and ask your first question.
- [Connect your environment](https://docs.rubixkube.ai/getting-started/connect-your-environment.md): Install the RubixKube Observer on Kubernetes, AWS, GCP, or a Linux VM. The Observer discovers your topology and streams signals to RubixKube Cloud.
- [How RubixKube works](https://docs.rubixkube.ai/concepts/how-rubixkube-works.md): The short version: a mesh of specialised AI agents runs the OPEL loop (Observe, Plan, Execute, Learn) against your infrastructure, backed by a memory that compounds over time.
- [The OPEL loop](https://docs.rubixkube.ai/concepts/opel-loop.md): Observe, Plan, Execute, Learn. The four-phase rhythm that every RubixKube agent follows, policy-driven and memory-backed.
- [The Observer Agent](https://docs.rubixkube.ai/concepts/observer-agent.md): The Observer is the eyes of RubixKube. It runs near your workload, discovers topology, collects signals, and keeps the Knowledge Graph current.
- [The Memory Engine](https://docs.rubixkube.ai/concepts/memory-engine.md): Private, compounding memory for your infrastructure. Every incident, RCA, and correction stays in your workspace and sharpens the next investigation.
- [The Knowledge Graph](https://docs.rubixkube.ai/concepts/knowledge-graph.md): A live model of your infrastructure: every resource, relationship, signal, and incident in one queryable graph.
- [The Agent Mesh](https://docs.rubixkube.ai/concepts/agent-mesh.md): A coordinated set of specialised AI agents, each expert in one slice of reliability work. They share a Knowledge Graph and run the OPEL loop together.
- [Skills](https://docs.rubixkube.ai/concepts/skills.md): Structured runbooks the SRI Agent can invoke. Turn operational know-how into reusable workflows with the open Agent Skills format.
- [Insights](https://docs.rubixkube.ai/concepts/insights.md): An Insight is an anomaly RubixKube thinks a human should know about. Learned baselines, scoped severity, and evidence behind every card.
- [Root Cause Analysis](https://docs.rubixkube.ai/concepts/root-cause-analysis.md): Evidence-linked causal chains for every investigation. How the RCA Pipeline Agent builds them, what is inside, and how to read one well.
- [Actions](https://docs.rubixkube.ai/concepts/actions.md): The unit of change in RubixKube. Every recommended fix is an Action with an explicit scope, blast radius, and approval path.
- [Environments](https://docs.rubixkube.ai/concepts/environments.md): The unit of scope in RubixKube. A Kubernetes cluster, an AWS account, a GCP project, or a set of VMs. Mix any number in one workspace.
- [Safety and Guardrails](https://docs.rubixkube.ai/concepts/guardrails.md): How RubixKube stays autonomous without being reckless. Read-only by default, explicit approvals on every change, evidence-backed recommendations, and data handling you can actually audit.
- [Kubernetes with RubixKube](https://docs.rubixkube.ai/environments/kubernetes.md): Install the RubixKube Observer on any Kubernetes cluster. EKS, GKE, AKS, KIND, and bare metal clusters on v1.24 or later are all supported.
- [AWS with RubixKube](https://docs.rubixkube.ai/environments/aws.md): Account-level observation across EC2, RDS, Lambda, S3, ELB, CloudTrail, and CloudWatch. Installer creates the read-only IAM role and policy for you.
- [GCP with RubixKube](https://docs.rubixkube.ai/environments/gcp.md): Project-level observation across GCE, GKE, Cloud SQL, Cloud Run, Cloud Storage, and Cloud Functions. Installer creates the service account and attaches viewer roles.
- [Azure with RubixKube](https://docs.rubixkube.ai/environments/azure.md): AKS today through the Kubernetes path. Subscription-level Azure observation for App Service, Azure VMs, and managed data is on the roadmap.
- [Bare metal and Linux VMs with RubixKube](https://docs.rubixkube.ai/environments/vm.md): Host-level monitoring for any Linux VM or bare metal server. CPU, memory, disk, network, process, and systemd signals.
- [Dashboard](https://docs.rubixkube.ai/using/dashboard.md): Your command centre for Site Reliability Intelligence. Health, insights, actions, RCA reports and infrastructure at a glance across every environment.
- [Rubix Chat Agent](https://docs.rubixkube.ai/using/rubix-chat-agent.md): The conversational surface of RubixKube. Ask your infrastructure questions in plain English, run skills, and launch investigations from the terminal or the console.
- [Magic Insights](https://docs.rubixkube.ai/using/insights.md): Incident detection, root cause analysis, and evidence-based troubleshooting in the RubixKube console.
- [Action Center](https://docs.rubixkube.ai/using/action.md): Manage and track remediation actions from RCA reports and incidents in one place.
- [RCA Reports](https://docs.rubixkube.ai/using/rca-reports.md): Evidence-linked root cause reports for every incident. Observed conditions, causal chain, recommended actions, and verification, all in one place.
- [Reliability Intelligence](https://docs.rubixkube.ai/using/analytics.md): Actionable reliability metrics for engineering leadership. Time saved, MTTU, RCA coverage, risk concentration, recurrence, and resolution velocity across every environment.
- [Environments](https://docs.rubixkube.ai/using/environments.md): Understand and operate Kubernetes, AWS, GCP, and VM environments in one RubixKube workspace.
- [Infrastructure Topology](https://docs.rubixkube.ai/using/infrastructure.md): Visualise and navigate your infrastructure with interactive topology views across Kubernetes, AWS, GCP, and VMs.
- [Integrations](https://docs.rubixkube.ai/using/integrations.md): Connect RubixKube with Slack, PagerDuty, Linear, Notion, and your own private tools.
- [Skills](https://docs.rubixkube.ai/using/skills.md): Enable system skills or create custom ones to extend the capabilities of your agents.
- [Notifications](https://docs.rubixkube.ai/using/notifications.md): Route insights, RCA reports, and pending actions to the channels your team already uses. Slack, Microsoft Teams, PagerDuty, email.
- [Team and Workspace Management](https://docs.rubixkube.ai/using/workspace.md): Invite team members, manage roles, and collaborate on reliability inside a RubixKube workspace.
- [Agent Skills Store](https://docs.rubixkube.ai/using/agent-skills-store.md): Browse and enable system skills built by RubixKube, or create your own. Every skill is a reusable runbook the SRI Agent can run.
- [GitHub integration](https://docs.rubixkube.ai/integrations/github.md): Give RubixKube agents read and write access to your GitHub repos so they can correlate deployments with incidents, cite PRs in RCAs, and propose soft remediations as pull requests.
- [GitLab integration](https://docs.rubixkube.ai/integrations/gitlab.md): Give RubixKube agents read and write access to your GitLab repos so they can correlate CI deployments with incidents, cite MRs in RCAs, and propose soft remediations as merge requests.
- [Notion integration](https://docs.rubixkube.ai/integrations/notion.md): Connect your Notion workspace so RubixKube agents can read team docs, publish RCAs as pages, and keep post-mortems in the wiki you already use.
- [Atlassian Rovo integration](https://docs.rubixkube.ai/integrations/atlassian-rovo.md): Connect the Atlassian stack so RubixKube agents can create Jira tickets, publish Confluence pages, read Bitbucket deploys, and bridge with Rovo agents.
- [Linear integration](https://docs.rubixkube.ai/integrations/linear.md): Connect Linear so RubixKube agents can read your projects, create tickets from insights and RCAs, and keep incident state in sync with the tool your team already tracks work in.
- [PagerDuty integration](https://docs.rubixkube.ai/integrations/pagerduty.md): Connect PagerDuty so RubixKube agents can trigger incidents for critical insights, sync acknowledgement and resolution, and keep the on-call picture in step with RubixKube.
- [Rubix in Slack](https://docs.rubixkube.ai/integrations/slack.md): Rubix, the RubixKube agent, in your Slack threads: mention it in a channel or message it directly, approve its changes with buttons, and see incidents as cards.
- [Microsoft Teams integration](https://docs.rubixkube.ai/integrations/microsoft-teams.md): Teams is both an interface to RubixKube and a notification channel. Talk to the bot in any channel to ask questions, trigger investigations, and approve actions. Get insights, RCAs, approvals, and digests where your team already lives.
- [New Relic integration](https://docs.rubixkube.ai/integrations/new-relic.md): Connect New Relic to RubixKube today via the custom integrations path. Agents can run NRQL, read APM service health, and pull alert state to enrich investigations and RCAs.
- [Datadog integration](https://docs.rubixkube.ai/integrations/datadog.md): Connect Datadog to RubixKube today via the custom integrations path. Agents can query metrics, logs, monitors, and APM services to enrich investigations and RCAs.
- [Dynatrace integration](https://docs.rubixkube.ai/integrations/dynatrace.md): Connect Dynatrace to RubixKube today via the custom integrations path. Agents can pull problems, entity health, and metrics to enrich investigations and RCAs.
- [Prometheus integration](https://docs.rubixkube.ai/integrations/prometheus.md): Connect Prometheus to RubixKube today via the custom integrations path. Agents can run PromQL and pull Alertmanager state to enrich investigations and RCAs.
- [Grafana integration](https://docs.rubixkube.ai/integrations/grafana.md): Connect Grafana to RubixKube today via the custom integrations path. Agents can read dashboards, alert state, and datasource queries to enrich investigations and RCAs.
- [Sentry integration](https://docs.rubixkube.ai/integrations/sentry.md): Connect Sentry to RubixKube today via the custom integrations path. Agents can read issues, events, and releases to enrich investigations and RCAs.
- [Custom MCP Servers](https://docs.rubixkube.ai/integrations/custom-mcp-servers.md): Extend RubixKube with your own Model Context Protocol servers. Expose internal tools to the SRI Agent with typed arguments and scoped authorisation.
- [Custom REST Integrations](https://docs.rubixkube.ai/integrations/custom-rest.md): Point RubixKube at a REST API with an OpenAPI spec. Every endpoint becomes a tool the SRI Agent can call, with scoped auth and approval gates.
- [OAuth and authentication types](https://docs.rubixkube.ai/integrations/oauth-and-auth-types.md): Reference for every authentication method RubixKube supports on integrations. API keys, OAuth 2.0, mTLS, signed requests, and more.
- [How to monitor infrastructure health with RubixKube](https://docs.rubixkube.ai/tutorials/monitor-infrastructure-health.md): Set up continuous health monitoring across Kubernetes, AWS, GCP, and VMs. Learn what each dashboard surface tells you and when to act.
- [How to automate incident remediation with RubixKube](https://docs.rubixkube.ai/tutorials/automate-incident-remediation.md): Go from detection to safe resolution. Learn how RubixKube surfaces the right RCA, recommends the fix, and how approvals keep you in control.
- [Talk to your infrastructure](https://docs.rubixkube.ai/tutorials/talk-to-infra.md): Ask your infrastructure questions in plain English. Chat is RubixKube's conversational gateway, backed by the SRI Agent and the whole Agent Mesh.
- [Chat Basics: Your First Queries](https://docs.rubixkube.ai/tutorials/chat-basics.md): Learn the Chat interface with hands-on examples. Start here for your first conversation with RubixKube.
- [Troubleshooting with Chat](https://docs.rubixkube.ai/tutorials/chat-troubleshooting.md): Use Chat to investigate real incidents: OOMKilled, CrashLoop, ImagePullBackOff, latency spikes.
- [Advanced Chat: Personas and Workflows](https://docs.rubixkube.ai/tutorials/chat-advanced.md): Master Chat for different roles. SRE, DevOps, and platform engineer workflows with real examples.
- [Cost analysis with Chat](https://docs.rubixkube.ai/tutorials/chat-cost-analysis.md): Analyse and reduce infrastructure cost using RubixKube Chat. Rightsizing, anomaly hunting, weekly budget reviews.
- [How to add custom agent skills](https://docs.rubixkube.ai/tutorials/add-custom-agent-skills.md): Turn a runbook into a reusable skill the SRI Agent can run. This tutorial covers the skill file format, allowed tools, scope, and testing.
- [How to add custom integrations](https://docs.rubixkube.ai/tutorials/add-custom-integrations.md): Connect RubixKube to your own internal tools. Covers Custom MCP servers, REST integrations, and the OAuth flows that make them safe.
- [Troubleshooting Guide](https://docs.rubixkube.ai/support/troubleshooting.md): Diagnose and resolve common issues in RubixKube.
- [Frequently Asked Questions](https://docs.rubixkube.ai/support/faq.md): Quick answers to common RubixKube questions.
- [System Status](https://docs.rubixkube.ai/support/system-status.md): Check the real-time health of all RubixKube platform services at rubixkube.ai/status. See current operational state, 7-day uptime, and 30-day incident history.
- [Glossary](https://docs.rubixkube.ai/support/glossary.md): Canonical definitions for every RubixKube term: agents, environments, RCAs, skills, and the rest of the vocabulary.
- [Rubix CLI overview](https://docs.rubixkube.ai/cli/overview.md): The Rubix CLI brings RubixKube's Site Reliability Intelligence to your terminal. Chat with infrastructure, investigate incidents, and run skills without leaving the command line.
- [Install the Rubix CLI](https://docs.rubixkube.ai/cli/installation.md): Install the Rubix CLI globally with npm, or run it without installing via npx. Node.js 18 or later required.
- [Using the Rubix CLI](https://docs.rubixkube.ai/cli/usage.md): Launch, log in, start a chat, resume sessions, and drive Rubix from your terminal. Keyboard-first workflows for SREs and platform engineers.
- [Rubix CLI configuration](https://docs.rubixkube.ai/cli/configuration.md): Where credentials and preferences live, how per-folder trust works, and how to switch environments and models.
- [Rubix CLI command reference](https://docs.rubixkube.ai/cli/commands.md): Every top-level command and every slash command in the Rubix CLI. Generated from the CLI source so it stays accurate.
- [Rubix CLI examples](https://docs.rubixkube.ai/cli/examples.md): Concrete flows for daily on-call, post-deploy verification, cost reviews, and CI automation using the Rubix CLI.
- [Rubix CLI troubleshooting](https://docs.rubixkube.ai/cli/troubleshooting.md): Common issues with the Rubix CLI: login failures, stale credentials, proxy and firewall problems, Node.js version mismatches.
- [Rubix CLI FAQ](https://docs.rubixkube.ai/cli/faq.md): Quick answers to common questions about the Rubix CLI: scope, safety, CI usage, Windows support, and more.
- [Kepler overview](https://docs.rubixkube.ai/kepler/overview.md): Kepler is RubixKube's desktop AI agent for SREs and on-call operators. Chat with your infrastructure, run read-only checks, drive a real terminal and browser, and keep a memory of every incident, all on your own machine.
- [Install Kepler](https://docs.rubixkube.ai/kepler/installation.md): Download and install Kepler on macOS, Windows, or Linux. System requirements, first launch, automatic updates, and how to uninstall.
- [Kepler quickstart](https://docs.rubixkube.ai/kepler/quickstart.md): Go from a fresh install to your first approved change in ten minutes. Sign in, connect a model, open a workspace, ask a read-only question, then make one small change in Assist.
- [The Kepler workspace](https://docs.rubixkube.ai/kepler/workspace.md): A tour of the Kepler window: the sidebar, the chat, tabs and split panes, and the Terminal, Browser, Editor, Memory, Feed and Topology panels. Learn the shortcuts and how workspaces and sessions fit together.
- [Remote access: work on other machines](https://docs.rubixkube.ai/kepler/remote-access.md): Point a Kepler thread at a server, a cloud VM or a sandbox instead of your laptop. Kepler uses your own logins, asks before it uses a machine, shows which machine every command ran on, and lets you cap what it may do on each one.
- [Postures: Observe, Assist, and Yolo](https://docs.rubixkube.ai/kepler/postures.md): How much Kepler does on its own. Observe is read-only, Assist asks before every change, Yolo acts within limits. How to switch, what the approval dock asks, and what each answer remembers.
- [Permissions and the always-blocked list](https://docs.rubixkube.ai/kepler/permissions.md): Per-tool allow, ask and block for files, shell and MCP servers, the command rules you teach Kepler, and the always-blocked list of destructive commands that no setting can unlock.
- [Memory and kepler.md](https://docs.rubixkube.ai/kepler/memory.md): How Kepler remembers your services, incidents, conventions and runbooks, the three files you own (kepler.md, pref.md, user.md), the Memory panel, and incognito sessions.
- [Topology: the map of your infrastructure](https://docs.rubixkube.ai/kepler/topology.md): Kepler keeps a map of your clusters, hosts, cloud accounts, databases and repositories, read from the config and CLIs already on your machine. Learn what it finds, how to add what no scan can see, and how to answer what Kepler proposes.
- [Extend Kepler: integrations, skills, plugins and specialists](https://docs.rubixkube.ai/kepler/extend.md): Connect Kepler to your stack with MCP integrations, keep credentials in the system keychain, add skills, custom slash commands, hooks and plugins, and save specialist agents for questions you ask often.
- [Customize: skills, commands, agents, hooks and plugins in one place](https://docs.rubixkube.ai/kepler/customize.md): Settings > Customize is where you manage everything that changes how Kepler works: plugins, skills, hooks, slash commands and specialist agents. One search, one list per kind, and a workspace filter.
- [Hooks: your own checks in Kepler's work](https://docs.rubixkube.ai/kepler/hooks.md): Run your own script, HTTP call, MCP tool or model check at moments in Kepler's work: before a tool runs, when a message arrives, when a run ends, when a watcher fires. Hooks can block or ask. They never approve.
- [Plugins and marketplaces](https://docs.rubixkube.ai/kepler/plugins.md): Install bundles of skills, slash commands, agents, hooks, MCP servers and chat channels in one step, from a marketplace, GitHub, a git URL or a folder. See everything a plugin runs before it runs.
- [The decision model](https://docs.rubixkube.ai/kepler/decision-model.md): An optional second model that answers typed questions in about a second. Kepler uses it to double-check commands and MCP calls before they run, and you can use it in chat, war rooms and automation steps to classify, score and pick.
- [Everyday workflows with Kepler](https://docs.rubixkube.ai/kepler/workflows.md): Prompt recipes for the work SREs do most in Kepler: triage an alert, investigate a pod, write an RCA, review a deploy, read a dashboard, bring terminal output into the chat, write a runbook, hand off on-call.
- [Automations: checks and briefs that run without you](https://docs.rubixkube.ai/kepler/automations.md): Build an automation from a plain description: a trigger (a schedule, a webhook, a check that changes, or a /command in chat) and a list of steps. Kepler runs it without you in the room and only changes things you said it may.
- [Run a war room](https://docs.rubixkube.ai/kepler/war-room.md): Kepler's incident mode: Kepler leads, spawns agents that each work one angle of the incident in their own pane, and gathers approvals and findings in one place.
- [Watchers: ask Kepler to keep an eye on something](https://docs.rubixkube.ai/kepler/watchers.md): A watcher is a read-only check Kepler runs in the background on a schedule. When it fires, Kepler wakes on its own, looks into it, and comes back with a suggestion.
- [Chat with Kepler from your phone](https://docs.rubixkube.ai/kepler/channels.md): Connect WhatsApp, Slack, Telegram, Discord or a channel from a plugin, and carry on the same session from your phone: ask questions, answer approvals, run automations, get woken by watchers while you are away.
- [Slash commands and keyboard shortcuts](https://docs.rubixkube.ai/kepler/commands.md): Every built-in Kepler slash command, how to write your own in .kepler/commands, how skills, plugin commands and automations show up in the / menu, and the full keyboard shortcut table for macOS, Windows and Linux.
- [Kepler settings reference](https://docs.rubixkube.ai/kepler/settings.md): Every Kepler setting, section by section: General, Appearance, Plan & Usage, Models, Tools & MCP, Customize, Permissions, Secrets, Gateway, Remote access, Memory, Watchers, Topology and Experiments.
- [Troubleshoot Kepler](https://docs.rubixkube.ai/kepler/troubleshooting.md): Fixes for the problems Kepler users hit most: the agent never connects, a model key is rejected, an integration will not sign in, a command is blocked, a folder is not visible, a panel is blank, memory looks stale, a watcher never fires, a channel drops.
- [Kepler FAQ](https://docs.rubixkube.ai/kepler/faq.md): Answers to common questions about Kepler: getting started, safety and approvals, models and cost, data and privacy, and platform support.
- [Kepler changelog](https://docs.rubixkube.ai/kepler/changelog.md): What is new in each Kepler release: MCP sign-in and war room fixes in 0.1.10-beta, hooks, plugins and chat-triggered automations in 0.1.9-beta, the decision model in 0.1.8-beta, remote access and a richer topology map in 0.1.7-beta, tabs and split panes in 0.1.6-beta, and automations in 0.1.5-beta.

## Optional

- [Blog](https://rubixkube.ai/blog)
- [Console](https://console.rubixkube.ai)
- [Status](https://rubixkube.ai/status)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.