Track 3 Tier 1: API how-tos (GetResourceReservations, UndeployApps, RemoveEntitiesFromTopology) + resolution diagnostics

[nahumtimerman]

Fills the highest-priority documentation gaps from the how-to coverage audit: three shipped CloudShell Automation API methods that had no usage documentation, plus the abstract-resolution failure diagnostics that shipped undocumented.

New articles (devguide/available-cs-api/useful-cs-api-examples/)

• Getting a Resource's Reservations — GetResourceReservations(resourceFullPath, includeHistorical, maxResults); active/pending/completed, pagination (TotalCount vs page), Python example.
• Undeploying Apps While Keeping the Resource — UndeployApps(resourcesFullPath[]); destroys the cloud-provider VM but keeps the resource (vs delete), bulk-teardown pattern.
• Removing Entities from a Topology — RemoveEntitiesFromTopology(topologyName, entityIds[]); removes services/abstract resources (+ attached routes) from a blueprint.

Updated

• Reserving blueprints fails on nonexistent resource conflicts — new "Reading the resolution diagnostics" section explaining the failure notification's View Details breakdown (Unresolvable / Conflict-blocked / Resolved / Routes), correctly scoped: it appears only for truly-unresolvable requirements; conflict-only failures still show the conflict-resolver dialog.

Grounding & validation

• Every signature, parameter, default, status value, and behavior is verified against the Trunk server code and integration test scenarios (e.g. RestGetResourceReservationsScenarios, RemoveEntitiesFromTopologyScenarios), then adversarially re-checked. Notably UndeployApps takes only resourcesFullPath[] (no reservation id).
• npm run build passes (only the 3 pre-existing abstract-resources.md broken links remain). Added undeploy/Undeploying/undeploys + pagination terms to .wordlist.txt.

🤖 Generated with Claude Code