Build operational context in a RunWhen workspace by adding Rules, Commands, and Knowledge Base articles. Use when: (1) The user wants to improve assistant investigation quality or tune workspace behavior, (2) Onboarding engineers or teams to a workspace, (3) Adding institutional knowledge, architecture docs, or…
Build, test, and commit a RunWhen automation task (SLX). Use when: (1) Creating a new health check or monitoring task, (2) Building a troubleshooting runbook or diagnostic script, (3) Writing and testing a bash or python script for RunWhen runners, (4) Committing an SLX with commitslx, (5) Running runscriptandwait or…
Render a tested tool-builder task as a private Custom Discovery CodeCollection for GitOps. Use when: (1) The user wants version-controlled storage instead of commitslx to a workspace, (2) Publishing tool-builder output to a private git repo shaped like simple-private-codecollection, (3) After runscriptandwait succeeds…
Set the hierarchy on an SLX to define UI grouping and path construction from tags. Use when: (1) Committing an SLX and need to set its hierarchy parameter, (2) Organizing SLXs into groups for the map view, (3) The user asks about hierarchy, SLX grouping, additionalContext, or map view organization, or (4) Configuring…
Set the resourcePath on an SLX for workspace-chat indexing and search. Use when: (1) Committing an SLX and need to set its resourcepath parameter, (2) Configuring SLX metadata for search and indexing, (3) The user asks about resourcePath, additionalContext, or how SLXs are found by workspace chat, or (4) Improving SLX…
Create and tailor a new AI Assistant (Agent / persona) in a RunWhen workspace. Use when: (1) The user wants a dedicated assistant for a tech stack, team, or domain (e.g. 'Azure DevOps', 'Postgres on-call'), (2) Creating, updating, or deleting an assistant via createassistant / updateassistant / deleteassistant, (3)…
Discover and select runner locations for RunWhen scripts. Use when: (1) Choosing where to execute a task or script, (2) The user asks about runner locations or execution environments, (3) Multiple locations exist and you need to pick the right one, or (4) Building a task that needs a specific cluster or region for…
Discover and configure secrets for RunWhen SLX scripts. Use when: (1) Choosing which secrets to map for a task via secretvars, (2) The user asks about secret configuration, authentication, or credentials, (3) Building a task that needs API tokens, kubeconfigs, or service account keys, or (4) Understanding how secrets…
Search the RunWhen CodeBundle Registry for production-ready automation and deploy it as an SLX. Use when: (1) Adding monitoring or health checks for Kubernetes, databases, or cloud services, (2) The user asks to create a health check and a registry codebundle may already exist, (3) Deploying pre-built automation…
Create, update, and manage workspace chat commands (slash-commands) that package reusable investigation procedures. Use when: (1) The user wants to create repeatable diagnostic workflows or slash-commands, (2) Building onboarding flows or standardized investigation patterns, (3) Creating or editing commands via…
Create, update, and manage Knowledge Base articles that provide operational context to AI assistants. Use when: (1) The user wants to add institutional knowledge, architecture notes, or ownership details, (2) Creating or updating KB articles via createknowledgebasearticle or updateknowledgebasearticle, (3) Adding…
Create, update, and manage workspace chat rules that shape how AI assistants interpret infrastructure findings. Use when: (1) The user wants to suppress noise or adjust investigation priorities, (2) Adding operational context or tuning assistant behavior, (3) Creating or editing chat rules via createchatrule or…
Run an existing SLX (health check, task, or automation) that is already committed to a workspace. Use when: (1) The user asks to execute, trigger, or run an existing task, (2) Running a health check or diagnostic on demand, (3) Calling runslx to execute a previously committed SLX, or (4) The user references a specific…
Validate a RunWhen MCP server installation is working end-to-end. Use when: (1) First connecting to a new MCP server, (2) After deployment, upgrade, or config change, (3) The user asks to verify, validate, check, or test the MCP setup, (4) Troubleshooting connectivity or permission issues, or (5) The user says…