@calltelemetry/ct-lab-mcp weekly downloads statistics

@calltelemetry/ct-lab-mcp

Model Context Protocol (MCP) Server for Unified Call Telemetry Lab, Fleet, and Appliance Management across Proxmox VE (ESXi is retired)

@calltelemetry/ct-lab-mcp is the single public MCP façade for Call Telemetry lab operations. It exposes high-level hypervisor management (Proxmox VE only; ESXi was retired 2026-10-15), disposable VM workflows, and secure appliance diagnostics while delegating canonical asset identity, lease ownership, and release-line policy to the private lab-inventory CLI supplied by ct-meta.

---

Architecture

AI Clients (Claude Desktop / Cursor / Web / AI Agents)
                  │
                  ▼
       [MCP Server / Transports]
       ├── Stdio Transport (stdout JSON-RPC, stderr SafeLogger)
       └── Express SSE Transport (/sse, /messages, /health)
                  │
                  ▼
         [MCP Tool Registry]
  ├── list_hosts          (Proxmox capacity: live free RAM/CPU/disk, health; 8GB unit estimates are deprecated)
  ├── list_inventory      (Unified VM fleet inventory with Linear tags)
  ├── claim_slot          (DEPRECATED alias -> create_appliance_claim; slot retirement ADR (REL-1746))
  ├── release_slot        (DEPRECATED alias -> delete_appliance_claim)
  ├── find_slot / list_slots (DEPRECATED aliases -> list_hosts / list_appliance_claims)
  ├── purge_slots         (RETIRED: explicit SLOT_VERB_RETIRED error; use list_expired_vms / cleanup_expired_vms)
  ├── provision_vm        (Disposable VM provisioning on Proxmox)
  ├── proxmox_quick_provision (Size-driven Proxmox fleet plan/apply; host=auto ranks by free RAM then disk)
  ├── pin_pool_slots      (RETIRED: explicit SLOT_VERB_RETIRED error; pass webImage to proxmox_quick_provision)
  ├── vm_action           (Power lifecycle with non-bypassable safety guardrails)
  ├── host_maintenance_shutdown (Guarded whole-host drain and graceful shutdown)
  ├── lab_mission         (Canonical manifest-bound multi-phase mission)
  ├── lab_dev             (Canonical leased remote-development lifecycle)
  ├── list_appliance_claims (Kubernetes ApplianceClaim CRD fleet discovery)
  ├── create_appliance_claim (Declarative CRD provisioning; profile=demo|bare|bundle; demo/bundle allocate node, vmid, vmName and targetIp from the canonical lab inventory and return `allocation`; bundle installs the exact release bundle with install-exact-release.sh, no OVA; bare tracks an existing guest with explicit vmid+vmName; node defaults to auto: least-loaded eligible Proxmox host, recorded in calltelemetry.com/placement)
  ├── list_held_claims    (Read-only: claims the operator holds instead of destroying, from its audit report)
  ├── lab_capacity        (Read-only: per node free RAM, the reserve and the claims holding RAM, grouped active/expired/paused/held/deleting/wedged/fixture)
  ├── release_appliance_claim_without_destroy (Ask the operator to drop a claim and leave the guest; honoured only for a template, absent guest or guest-less claim)
  ├── renew_appliance_claim (Extend a claim TTL; never shortens)
  ├── claimed_import_qcow (Image to claimed guest: claim-first bare claim, then canonical vm provision --source published-qcow)
  ├── delete_appliance_claim (Operator teardown trigger via CRD finalizer)
  ├── sideload_appliance_runner (Isolated runner commands & container overrides)
  ├── get_appliance_id    (Hardware UUID -> CT-APP-XXXX-XXXX-XXXX derivation)
  ├── sideload_license    (Standardized calltelemetry-license-v2 envelope)
  ├── verify_appliance_demo (Automated 5-point demo verification gate)
  ├── appliance_update    (Plan/apply a release update of a claimed guest via canonical lab-inventory vm update)
  ├── uat_credential      (Executor-mediated UAT credential injection: claim-authorized grant of an opaque single-use handle; redeem runs the executor-configured UAT runner with the credential in the child env only; the value is never returned)
  └── appliance_exec      (RETIRED: refuses with CANONICAL_PHYSICAL_SLOT_AUTHORITY_REQUIRED; see appliance_update, uat_access)
                  │
  ┌───────────────┼───────────────┬───────────────────┐
  ▼               ▼               ▼                   ▼
[MultiHypervisor] [Slot Engine]   [Safety Guardrails] [Appliance SSH]
  └── Proxmox VE   ├── ProxmoxLab  ├── Proxmox fixtures ├── ssh2 port 2222
      (lab, mini,  │   (2x8GB max) │   (100-104, 9000s) ├── cli.sh / db vacuum
       ibm64)      ├── ProxmoxMini └── CUCM, CUBE,      ├── migrate / logs
                   │   (4x8GB max)     Runners, K8s     └── Redaction Engine
                   └── ProxmoxIbm64
                          │
                          ▼
            [Lab Credential Resolver]
            ├── Tier 1: Process Environment (LAB_PROXMOX_*, LAB_VM_*)
            ├── Tier 2: Doppler REST API v3 (DOPPLER_TOKEN)
            └── Tier 3: In-Memory Mock Fallback / Offline Dev

In a governed ct-meta installation there is only one registered MCP server: @calltelemetry/ct-lab-mcp. The Python lab-inventory program remains a private canonical policy engine behind the Node façade; it is not registered as a second public MCP. Set CT_LAB_CANONICAL_REQUIRED=1 to fail closed rather than use the legacy Node VMID/name slot state when the bridge is unavailable.

---

Key Features…