Uptime: Wiki | Prospect

Parameters

Path Parameters

id*integer

Entity id

Formatint64

Range0 <= value

Code example

curl -X GET "https://example.com/api/v1/prospects/0"

Response example

{  "id": 0,  "name": "string",  "archetype": "string",  "tier": "string",  "signup_bonus_dollars": 0.1,  "compliance_required": "string",  "required_arch": "string",  "requires_multi_az": true,  "required_region": 0,  "services_summary": "string",  "service_kinds": [    "string"  ],  "estimated_monthly_revenue_dollars": 0.1,  "expires_at_ns": 0,  "initial_vm_count": 0,  "vcpu_per_vm": 0,  "memory_mb_per_vm": 0,  "block_gb_per_vm": 0,  "growth_ceiling_vms": 0,  "fit_vcpu_after_pct": 0.1,  "fit_mem_after_pct": 0.1,  "fit_ssd_after_pct": 0.1,  "fit_can_serve_now": true,  "fit_oversubscribed": true,  "fit_over_capacity": true,  "fit_recommendation": "string",  "requires_gpu": true,  "gpu_slices_demand": 0,  "services": [    {      "class_tag": "string",      "label": "string",      "requested_count": 0,      "vcpu": 0,      "memory_mb": 0,      "storage_gb": 0,      "throughput_label": "string",      "durability": "string",      "vgpu_slices": 0,      "gpu_mem_mb": 0    }  ],  "demands_audit_logs": true,  "demands_air_gap": true,  "requires_private_segment": true,  "requires_dedicated_hosts": true,  "requires_fault_isolation": true,  "covering_segment_id": 0,  "covering_segment_name": "string",  "requires_uncontended": true,  "requires_dedicated": true,  "uncontended_hosts_available": 0}

Response body

200The entity

application/json

id*integer

Formatint64

Range0 <= value

name*string

archetype*string

tier*string

signup_bonus_dollars*number

Formatdouble

compliance_required*string

required_arch*string

Raw CPU-arch requirement tag ("x86_64" / "arm64" / …) this prospect's VM or K8s workload demands, or "" if arch-agnostic. A hard placement gate: the card voices it (jargon "x86-64 hosts" vs plain "Intel/AMD hosts") beside the compliance/air-gap chips. Godot maps the tag via the arch_req_<tag> glossary key.

requires_multi_az*boolean

required_region?|

Formatint32

Range0 <= value

services_summary*string

service_kinds?array<string>

Short-code service mix mirroring CustomerView.service_kinds (subset of {"fn","vm","obj","lb","db","cdn","k8s"}). The ops-console customer/prospect table renders these as upper-case chips, so a prospect row reads the same as an active-customer row.

estimated_monthly_revenue_dollars?number

List-price monthly run-rate this prospect would bill at signup, projected from its VM workload (count × per-VM hourly × hours/month). 0 for non-VM-shaped prospects. Lets the customers lens show a prospect's potential Revenue/hr before you sign them.

Formatdouble

expires_at_ns*integer

Formatint64

Range0 <= value

initial_vm_count?integer

Initial VM count this prospect will request at signup. 0 when the prospect doesn't run a VM workload (Functions / ObjectStore / etc.) — the renderer surfaces those via services_summary.

Formatint32

Range0 <= value

vcpu_per_vm?integer

Per-VM vCPU count. 0 when no VM workload.

Formatint32

Range0 <= value

memory_mb_per_vm?integer

Per-VM memory (MB). 0 when no VM workload.

Formatint32

Range0 <= value

block_gb_per_vm?integer

Per-VM block-volume size (GB). 0 when the ask carries no volume (legacy templates / non-VM workloads). Billed via BlockPricing alongside the VM's compute charge.

Formatint32

Range0 <= value

growth_ceiling_vms?integer

Growth ceiling (max VMs the customer can compound to) read from the archetype's growth.plateau_at_allocation. 0 = no archetype match or uncapped (Anchor / Enterprise). Renderer shows "may grow to ~N" when > initial_vm_count.

Formatint32

Range0 <= value

fit_vcpu_after_pct?number

Capacity-fit verdict (P2 in tasks/consequences.md) — projected fleet VM utilization after accepting this prospect's initial demand, on each axis. 1.0 = exactly full; > 1.0 = would overcommit. Lets the prospect card answer "can I actually serve this?" before signing. 0.0 for prospects with no VM workload.

Formatdouble

fit_mem_after_pct?number

Formatdouble

fit_ssd_after_pct?number

The DISK axis, on the same basis as the two above. contract_footprint has always computed Resource::SsdGb (an object-storage bucket, a CDN origin, a DB's backup copies, a VM's block volume); the projection threw it away, so a card asking for 821 GB showed no disk meter at all and read "Comfortable" beside it. 0.0 when the bundle asks for no bytes.

Formatdouble

fit_can_serve_now?boolean

True when the current fleet has free vCPU, memory and disk headroom for this prospect's initial demand right now (no new hardware needed).

fit_oversubscribed?boolean

True when accepting would push the fleet PAST physical into overcommit: the initial VMs still fit under the host overcommit ceiling, but CPU is shared and tenants may slow under load. Never a false "can't fit".

fit_over_capacity?boolean

True only for a genuine over-capacity fit: a VM-load prospect that won't fit even under overcommit (the "Over capacity, add hardware first" verdict). Boolean twin of that string so the card's Accept-anyway confirm gate keys off a field, not exact prose. False for no-VM / oversubscribed.

fit_recommendation?string

Engine-classified one-liner for the card: "Comfortable" / "Tight fit" / "Will run oversubscribed (CPU shared under load)" / "Over capacity, add hardware first" / "Needs GPU hardware — add a GPU host first" / "No VM load".

requires_gpu?boolean

True when any of this prospect's VM services demands a GPU (vGPU partition / required_accelerator == Gpu). Drives the card's GPU meter and makes the serve verdict GPU-aware — a GPU tenant on a GPU-less fleet reads "Needs GPU hardware", not a false "Comfortable".

gpu_slices_demand?integer

Total vGPU slices this prospect's initial VMs would reserve (initial_vm_count × per-VM slices). 0 for non-GPU prospects. The fleet-fit input the serve verdict weighs against free vGPU slices.

Formatint32

Range0 <= value

services?array<>

Per-service detail for the WHOLE bundle — one entry per service the prospect would sign for. The card renders each as a magnitude line so non-VM prospects (object-store / DB / functions / ...) read as more than a bare chip. Additive to the VM fields above, which still drive the capacity-fit bar.

demands_audit_logs?boolean

True when the prospect's archetype demands audit logging (a live SIEM + audit feed). Surfaced as a requirement chip so a PCI / government-contract prospect's otherwise-inexplicable rejection is legible ("Requires: audit logging").

demands_air_gap?boolean

True when the prospect's archetype demands an air-gapped facility.

requires_private_segment?boolean

True when the prospect wants a private/isolated network segment (VXLAN isolation) dedicated to its SLA tier. SOFT / recommended — unlike audit-logging / air-gap it does NOT hard-reject signing; it's satisfied by a private segment whose tier_rule matches this prospect's tier. Surfaced as a "Recommended" chip, distinct in tone from the hard requirement chips.

requires_dedicated_hosts?boolean

True when the prospect wants a private network on DEDICATED, homed hosts (single-tenant iron) — the 4× isolation tier (vs 2× for a plain private network). Implies requires_private_segment; the card shows one combined chip. msgpack-only (ProspectView is not carried in capnp).

requires_fault_isolation?boolean

True when the prospect's contract asks for fault isolation — copies kept off a shared fault domain (redundancy tier > 1 host replica, or an archetype requiring anti-affinity). Mirrors CustomerView.requires_fault_isolation so the prospect "Must clear" card and the accepted-customer Requirements row read the SAME key and cannot drift. msgpack-only (ProspectView is not carried in capnp).

covering_segment_id?integer

Id of the segment whose tier_rule already claims this prospect's tier — i.e. the segment the prospect would auto-join on signup via tier-rule precedence. 0 = none (no tier-rule segment covers this tier; the prospect would land on the Default shared segment). Lets the prospect card offer "isolate into " without a second round-trip. Mirrors the accept-path covering check.

Formatint32

Range0 <= value

covering_segment_name?string

Display name of the covering segment (empty when covering_segment_id is 0). Paired with the id so the card can label the existing-segment option without a client-side lookup.

requires_uncontended?boolean

True when the prospect's archetype contracts UNCONTENDED (1:1 CPU) — its workloads must land on hosts admitted to an uncontended segment (running 1:1, no overcommit). Wholly independent of dedication and anti-affinity spread. Drives the card's "Needs 1:1 CPU" chip so the player sees the guarantee (and the pool cost) before signing.

requires_dedicated?boolean

True when the prospect's archetype contracts DEDICATED, single-tenant hosts (private iron, no co-tenants). Mirrors requires_dedicated_hosts on the archetype's own flag; surfaced as a "Needs a dedicated host" chip. Tenancy guarantee, not a CPU one — distinct from uncontended.

uncontended_hosts_available?integer

Best-effort count of hosts the engine could convert into the shared uncontended (1:1) pool right now — open (unadmitted), serviceable, VM-serving hosts with physical room. Fleet-wide, not per-prospect. The card renders " host(s) available for the 1:1 pool" for an uncontended prospect so the player knows whether the pool can grow before signing. 0 = none convertible (would need new hardware).

Formatint32

Range0 <= value

Prospect

Parameters

Try the endpoint

Code example

Response example

Response body

On this page