[dofastted]

Summary

Fixes #732 by surfacing the existing vm.active / active.json state in the VM management UI and wiring delete flows to the existing activate API.

• Active badge: table, rack, detail header, and list detail card show Active when vm.active === true.
• Set as active: row context menu calls POST /api/panel/vms/:id/activate; ops tab button is disabled when already active.
• Delete active slot: if other schedulable, credential-valid slots exist, the confirm dialog offers 切换活跃并删除 (auto-activate the best candidate, then DELETE). If no safe candidate, delete is disabled with guidance to switch active manually first.

Out of scope (per issue)

• No backend auto-heal of active.json
• No admin API fallback for unhealthy active slots
• VERSION unchanged

Verification

• pnpm -C web test src/features/vm/active-vm.test.ts (4 tests)
• pnpm -C web exec tsc -b --noEmit

Operator upgrade

Node + web/dist only; no fetch:payload / wrap-cli sync required for this change alone.

  

---

Note

Low Risk
Frontend-only panel changes wiring to existing activate/delete APIs; delete safety is improved via client-side planning, not new server logic.

^{Reviewed by Cursor Bugbot for commit 7463a50. Configure here.}

[cursor]

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: ad1cab96-9386-4da7-8b4d-d4d2fc601cc1)

[cursor]

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: 8bc4dcd3-3ed9-4f27-bd87-6a9522998db8)