Skip to content
RunWhen

Guides

Guides are practical, outcome-focused walkthroughs for real workflows in RunWhen. They sit between reference docs and blog posts: opinionated enough to help you execute quickly, but detailed enough to support repeatable team practices.

Use this section when you want to do something end-to-end:

  • Configure assistant behavior for your environment
  • Standardize operational workflows across teams
  • Roll out repeatable triage and remediation patterns
  • Adopt advanced integrations and automation practices

  • MCP end-to-end flow (IDE to infrastructure)
    A diagram-led walkthrough of the Investigate and Build journeys over MCP — including how Build covers new automation, AI assistant rules and commands, and knowledge articles. Also explains how GitHub Copilot, Claude, and hosted vs self-hosted MCP fit together, in plain English.

  • RunWhen terms for AI audiences
    A short mapping from RunWhen vocabulary (Rules, Tasks, Commands, Knowledge) to familiar agent and LLM concepts — useful when explaining the product to teams who already think in tools, RAG, and system prompts.

  • Building Operational Context
    A walkthrough for adding Rules, Commands, and Knowledge to your workspace — moving from generic AI output to investigations your team trusts enough to act on.

  • Managing Context as Code (GitOps)
    The GitOps companion to Building Operational Context — keep Rules, Commands, Knowledge, and Custom Tasks in your own Git repository and reconcile them into a workspace from CI via the RunWhen API and MCP Tool Builder.

  • Bring Your Own LLM — overview

  • Self-Hosted LLM Configuration — Helm llmGateway / llmBootstrap for runwhen-platform

  • Azure OpenAI & AI Foundry — BYO LLM Setup
    Step-by-step setup for connecting RunWhen to your Azure OpenAI or Azure AI Foundry endpoint using a cross-tenant Service Principal trust model. Covers role assignments, endpoint configuration, and IP allowlisting for both endpoint types.

  • Kubernetes Kubeconfig Setup
    How to provide kubeconfigs to RunWhen — single file with multiple contexts, cloud-provider-generated kubeconfigs (AKS, GKE, EKS) and how they merge at runtime, and step-by-step service account and RBAC setup for self-managed clusters.

  • JFrog Artifactory / JCR Setup for RunWhen Self-Hosted
    Configure JFrog as the enterprise registry that fronts a self-hosted RunWhen Platform install. Covers Docker remotes, the virtual aggregator, optional Helm mirrors, the cluster pull secret, Helm values routing, and verification — with a data-flow diagram and a path-mapping reference.

  • Google Artifact Registry Setup for RunWhen Self-Hosted
    The native-GCP companion to the JFrog guide. Configure GAR remote repositories to proxy RunWhen images and third-party dependencies, wire a read-scoped GCP pull secret, and route Helm values per repo path — with a data-flow diagram and path-mapping reference.

  • OpenShift & OKD — Self-Hosted Installation
    Cluster-admin checklist for SCC exceptions, Vault RBAC, and image pull on OpenShift/OKD — what a namespace owner cannot do alone, with worked examples for bundled Postgres (Spilo) and pointers to the chart’s restricted vs exception value overlays.

  • Build a Custom Discovery CodeCollection
    Extend what RunWhen can see — Kubernetes CRDs, labeled workloads, cloud tags that aren’t in any published CodeCollection — by shipping a small private CodeCollection that carries only discovery rules and templates and delegates the runtime to the shared RunWhen Generic CodeCollection. Worked example: monitor Crossplane GCP Bucket CRDs with a fractional-conditions SLI and a condition-based TaskSet that files issues on Ready=False or Synced=False.

  • Author Custom Tasks with MCP
    Use Cursor or Claude Desktop to write, statically validate, and live-test a Python or Bash task against your dev workspace, then have the MCP agent render a version-controlled bundle into your private CodeCollection. Includes the round-trip check that keeps the base64-encoded GEN_CMD honest at PR review, the dev → stage → prod git-branch promotion model, and a security table pinning down where the LLM does and doesn’t touch infrastructure.

  • Common User Journeys — day-to-day usage flows by role and scenario
  • Live Demos — hands-on sandbox scenarios
  • Blog — engineering deep-dives, product updates, and field stories

What’s next

This library will continue to grow with integration guides, operational playbooks, and advanced configuration patterns. New guides will follow the same format: clear prerequisites, UI screenshots, specific examples, and measurable outcomes.