Skip to content

Use

This section covers the day-to-day usage of the RunWhen platform. Each area of the platform has a dedicated page with screenshots, step-by-step instructions, and common workflows.

If you want one end-to-end walkthrough for building operational context across Rules, Commands, and Knowledge, start with Guide: Building Operational Context.

Platform Navigation

The RunWhen workspace has three main areas, accessible from the left sidebar:

AreaIconPurpose
Workspace ChatChat bubbleNatural language troubleshooting with AI Assistants. Also shows Issues detected by background tasks.
Workspace StudioScissors/toolsManage Tasks, Rules, Commands, Knowledge, Workflows, and Assistants. This is where you configure what the platform knows and does.
Workspace ConfigurationGearWorkspace settings, user management, secrets, and Slack integration.

Pages in This Section

Workspace Chat

The primary troubleshooting interface. Describe a problem in natural language, and an AI Assistant investigates on your behalf — searching for existing issues, running diagnostic tasks, analyzing results, and recommending fixes.

Key topics: Starting a session, how the Assistant responds, scope filtering, selecting an assistant, chat history, prompting tips.

Workspace Studio

The management hub for everything the platform knows and can do. Contains six tabs: Tasks, Rules, Commands, Knowledge, Workflows, and Assistants.

Key topics: SLX management, creating tasks, defining rules and commands, managing knowledge sources, configuring workflows.

Related guide: Building Operational Context — where Rules, Commands, and Knowledge fit together, including durable workspace setup and short-term release/change overlays.

Engineering Assistants

AI-powered agents with configurable personas and permission levels. Each workspace comes with preset assistants (Eager Edgar, Cautious Cathy, Admin Abby, Dev Danica) and supports custom configurations.

Key topics: Assistant profiles, confidence thresholds, RBAC, Form vs YAML configuration, creating custom assistants.

Issues and Triage

Issues are findings raised by automated tasks that signal something in your environment needs attention. The Issues view provides a centralized list of all open findings, with tools to investigate, escalate, or ignore them.

Key topics: Issues list, issue detail, triage workflow, ignoring known issues, automatic resolution.

Common User Journeys

Step-by-step guides for typical workflows: investigating unhealthy services, responding to alerts, developer self-service troubleshooting, onboarding to a new workspace, setting up automated monitoring, and reviewing past incidents.

MCP Server

Connect your AI coding agent (Cursor, VS Code, Claude Desktop, Claude Code, or other MCP-capable CLIs) to RunWhen over the Model Context Protocol — locally or via RunWhen-hosted remote MCP on beta (https://mcp.beta.runwhen.com/mcp). Ask the AI assistant questions, build automation tasks, and commit SLXs — from your IDE or terminal.

Key topics: Installation (including Claude Desktop and CLI setup), local vs remote configuration, tool reference, Tool Builder workflow, script contract, RUNWHEN.md context.