> ## Documentation Index > Fetch the complete documentation index at: https://docs.kombify.io/llms.txt > Use this file to discover all available pages before exploring further. # Publish the server at a kombify.me address, or stop publishing it > Nothing is published by default: a paired server has no public route and the relay carries no traffic. The owner enables it per server with an explicit `service` name and `target_port` (the origin is always the Guard host's loopback, never another machine) and may withdraw it at any time. The relay authenticates the Guard, not the visitor, so what answers on that port must protect itself. Enabling requires the owner, the `techstack.servers.maintenance` entitlement and a fresh multi-factor sign-in (reason `server_reach_step_up_required`); withdrawing is never gated. Responds with the reach section of the server read model; the reconciler then creates or withdraws the route. ## OpenAPI ````yaml /techstack/reference/api/openapi.yaml put /api/v1/servers/{serverId}/reach openapi: 3.1.0 info: title: kombify-TechStack API description: > REST API for kombify-TechStack -- the hybrid cloud control plane for homelabs. kombify-TechStack manages infrastructure through a spec-driven architecture: validate YAML specs via CUE, generate OpenTofu configurations, and manage nodes through a gRPC agent protocol. ## Authentication Authenticated endpoints accept an OIDC/JWKS bearer token or an API key where documented. Self-hosted TechStack defaults to Pocket ID as identity provider. SaaS routes enter through the edge with Auth0 as the platform identity provider. PocketBase session exchange remains legacy migration compatibility only. ## Response Format **Success:** `{ "data": , "meta": { "total": int, "page": int, "per_page": int } }` **Error:** `{ "error": { "code": "", "message": "...", "details": } }` ## Pagination List endpoints accept `page` (default: 1) and `per_page` (default: 20, max: 100) query parameters. version: 1.0.0 contact: name: kombify url: https://kombify.io license: name: Apache-2.0 AND GPL-3.0 url: https://github.com/KombiverseLabs/kombify-TechStack/blob/main/LICENSE servers: - url: https://api.kombify.io/v1/techstack description: >- Production canonical (via Cloudflare edge gateway, strip-prefix). Clients MUST use this projection. The Gateway strips /v1/techstack and re-adds /api/v1 at the origin, so it resolves the /api/v1/* paths below cleanly. - url: http://localhost:5260 description: Local development - url: https://techstack.kombify.io description: Production (direct, no Gateway prefix) - url: https://api.kombify.io/v1/orchestrator description: >- Legacy StackKit-Orchestrator projection of the same origin. RETAINED for backward compatibility with existing consumers only. Do NOT use for mobile or new clients: combined with the /api/v1/* paths it double-resolves (yielding /v1/orchestrator/api/v1/...). Prefer /v1/techstack. security: - bearerAuth: [] tags: - name: Health description: Health checks and service info (no auth required) - name: Instance description: Local instance identity and enrollment metadata - name: Stacks description: Stack lifecycle management (create, provision, deploy, destroy) - name: Stacks - Specs description: Access persisted specification chain (intent, requirements, unified) - name: Stacks - Import/Export description: Import and export StackKits stack-spec.yaml configurations - name: Stacks - Drift description: Infrastructure drift detection and resolution - name: Unifier description: Spec validation, analysis, unification, and IaC generation pipeline - name: StackKits description: Infrastructure template catalog - name: Add-ons description: Add-on detection and registry - name: Workers description: Worker node registration and management - name: Servers description: Canonical server lifecycle, connection, and health state - name: Inventory description: >- Owner-scoped sanitized server and service read models for UI, API, and MCP clients - name: Pre-Checks description: Pre-deployment hardware and software validation - name: Agents description: Connected gRPC agent management and commands - name: Jobs description: Background job monitoring and SSE streaming - name: Discovery description: Network device discovery and SSH testing - name: Features description: Feature flags (read-only; managed centrally via the kombify Admin API) - name: Monthly Runtime description: Provider-free user API for subscription-backed stack runtime control - name: Runtime description: Recent runtime/deploy logs (job/deploy progress + container tail) - name: Wallet description: >- Redacted recovery-item wallet with biometric/platform-reauth-gated reveal. List responses strip secret/totp; reveal requires a fresh trusted-reauth HMAC proof (5-minute window) and is audited server-side. - name: Trust description: Pairing tokens for worker enrollment - name: Client Binding description: >- Short-lived HTTPS pairing for a second client and normal OIDC account handoff - name: Tunnel description: Cloudflare tunnel management for NAT traversal - name: Backups description: Database backup management (admin only) - name: Auth description: Authentication mode and SSO configuration - name: Metrics description: Prometheus-compatible instance metrics for admin/infra scrapers - name: Internal description: >- Internal SaaS-only APIs, including servicecall-protected executor contracts paths: /api/v1/servers/{serverId}/reach: put: tags: - Servers summary: Publish the server at a kombify.me address, or stop publishing it description: >- Nothing is published by default: a paired server has no public route and the relay carries no traffic. The owner enables it per server with an explicit `service` name and `target_port` (the origin is always the Guard host's loopback, never another machine) and may withdraw it at any time. The relay authenticates the Guard, not the visitor, so what answers on that port must protect itself. Enabling requires the owner, the `techstack.servers.maintenance` entitlement and a fresh multi-factor sign-in (reason `server_reach_step_up_required`); withdrawing is never gated. Responds with the reach section of the server read model; the reconciler then creates or withdraws the route. operationId: setServerPublicReach parameters: - $ref: '#/components/parameters/serverId' requestBody: required: true content: application/json: schema: $ref: '#/components/schemas/GuardPublicReachRequest' responses: '200': description: The reach section after the change content: application/json: schema: type: object properties: data: $ref: '#/components/schemas/GuardReach' meta: $ref: '#/components/schemas/ResponseMeta' '400': $ref: '#/components/responses/BadRequest' '401': $ref: '#/components/responses/Unauthorized' '403': $ref: '#/components/responses/Forbidden' '404': $ref: '#/components/responses/NotFound' '409': $ref: '#/components/responses/Conflict' '503': $ref: '#/components/responses/ServiceUnavailable' components: parameters: serverId: name: serverId in: path required: true description: Canonical server runtime ID schema: type: string schemas: GuardPublicReachRequest: type: object additionalProperties: false required: - enabled properties: enabled: type: boolean service: type: string pattern: ^[a-z]([a-z0-9-]{0,12}[a-z0-9])?$ description: >- Required when enabling. The route label; the kombify.me route is --. target_port: type: integer minimum: 1 maximum: 65535 description: >- Required when enabling. The port the service listens on, on the Guard host's loopback (the host is fixed). target_scheme: type: string enum: - http - https default: http GuardReach: type: object description: >- kombify Guard reach, persisted and never derived client-side: whether the server's kombify.me address answers through the Guard relay and whether the server is listed as a Connect resource. `ready` is shown as "Reachable and governed". A reach failure never blocks Guard enrollment or execution and never changes the server lifecycle. Absent on deployments without a Cloud connection. required: - state - kombify_me - connect - updated_at properties: state: type: string enum: - not_enrolled - pending - failed - ready - revoked - not_enabled description: >- pending is reach_pending, failed is reach_failed (see the reason codes of the two targets). not_enabled is a server the owner has not published at a kombify.me address (the default): no route exists and nothing is served. generation: type: integer format: int64 description: Binding generation, raised when the Guard key or pairing changes. public: type: object description: >- The owner's publication choice; nothing is published until it is enabled with a service and a port. required: - enabled properties: enabled: type: boolean service: type: string description: The route label the owner chose. scheme: type: string enum: - http - https target_port: type: integer minimum: 1 maximum: 65535 description: The port of the service on the Guard host's loopback. enabled_by: type: string enabled_at: type: string format: date-time kombify_me: $ref: '#/components/schemas/GuardReachTarget' connect: $ref: '#/components/schemas/GuardReachTarget' relay: type: object description: The Guard's own report of its relay connection. required: - state properties: state: type: string enum: - connecting - connected - denied - stopped reason_code: type: string reported_at: type: string format: date-time updated_at: type: string format: date-time ResponseMeta: type: object properties: total: type: integer page: type: integer per_page: type: integer managed_runtime_inventory_complete: type: boolean description: >- True only when the managed RuntimeLease authority was configured and read successfully for this inventory response. worker_inventory_sha256: type: string pattern: ^[0-9a-f]{64}$ description: >- Stable digest of the complete worker identity, attachment, freshness, and managed RuntimeLease authority snapshot used for pagination. managed_runtime_active_lease_ids: type: array items: type: string description: >- Canonically sorted IDs of visible managed RuntimeLeases classified as active for inventory and cleanup. This visibility projection is not capacity reservation or release authority. managed_runtime_inactive_lease_ids: type: array items: type: string description: >- Canonically sorted IDs of visible managed RuntimeLeases classified as cancelled or archived for inventory and cleanup. Inactive state does not prove provider absence or release managed-server capacity. managed_runtime_lease_generation_digests: type: array items: type: object required: - lease_id - resource_generation_digest properties: lease_id: type: string resource_generation_digest: type: string pattern: ^[0-9a-f]{64}$ description: >- Canonically sorted opaque bindings from each visible managed RuntimeLease ID to its exact provider-resource generation. Missing coverage makes destructive inventory proof invalid while leaving ordinary dashboard inventory readable. managed_runtime_duplicate_lease_ids: type: array items: type: string description: >- Managed RuntimeLease IDs represented by multiple persisted worker rows before UI projection deduplication; must be empty for destructive inventory proof. managed_runtime_attachment_conflict_lease_ids: type: array items: type: string description: >- Canonically sorted caller-visible managed RuntimeLease IDs with a conflicting tenant-wide worker attachment; must be empty for destructive inventory proof and never exposes foreign private lease or worker details. GuardReachTarget: type: object required: - state properties: state: type: string enum: - pending - registered - failed - revoked - not_enabled address: type: string description: kombify.me address (kombify_me target, registered only). node_id: type: string description: Connect techstack-node id (connect target, registered only). reason_code: type: string description: Stable structured reason for a pending or failed target. ErrorResponse: type: object required: - error properties: error: type: object required: - code - message properties: code: type: string enum: - INTERNAL_ERROR - NOT_FOUND - BAD_REQUEST - UNAUTHORIZED - FORBIDDEN - CONFLICT - VALIDATION_ERROR - RATE_LIMITED - SERVICE_UNAVAILABLE - METHOD_NOT_ALLOWED message: type: string details: description: Additional error context (validation errors, etc.) responses: BadRequest: description: Bad request content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' Unauthorized: description: Authentication required content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' Forbidden: description: Insufficient permissions content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' NotFound: description: Resource not found content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' Conflict: description: Request conflicts with existing state content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' ServiceUnavailable: description: Required control-plane persistence or dependency is unavailable content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' securitySchemes: bearerAuth: type: http scheme: bearer bearerFormat: JWT description: >- OIDC/JWKS bearer token. Self-hosted defaults to Pocket ID; SaaS uses Auth0 through the edge. PocketBase JWT is legacy migration compatibility only. ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.