> ## Documentation Index > Fetch the complete documentation index at: https://docs.gmicloud.ai/llms.txt > Use this file to discover all available pages before exploring further. # List Template Builds > Lists immutable Build attempts in reverse chronological order. ## OpenAPI ````yaml /api-spec/sandbox_api.yaml get /templates/{template_id}/builds openapi: 3.0.3 info: title: GMI Sandbox API description: >- REST API for GMI Sandbox: create isolated execution environments from templates, manage their lifecycle on the control plane, and run commands or transfer files through each sandbox's own data plane. contact: name: GMI Cloud Support email: support@gmicloud.ai version: '2.0' servers: - url: https://console.gmicloud.ai/api/v2 description: Control plane security: - bearerAuth: [] tags: - name: Sandbox description: | Sandbox instance lifecycle management (Sandbox v2, /api/v2/sandboxes) - name: SandboxTemplate description: Sandbox template management (Sandbox v2, `/api/v2/templates`) - name: Sandbox-Exec description: > Sandbox execution data plane (accessed via the sandbox host and sandbox access token). These endpoints are not under the control plane's `/api/v2` path; clients connect directly to `https://{sandbox_key}.{domain}` and authenticate with `X-Access-Token`. - name: Sandbox-Files description: > Sandbox file data plane (`/files`). **These endpoints are not under `/api/v2`** and do not use `Authorization: Bearer`: they are served by the sandbox's own data-plane entry point — clients connect directly to `https://{sandbox_key}.{domain}` and authenticate with `X-Access-Token`. `{sandbox_key}` is the `sandbox_key` returned by the create/connect endpoints, used for data-plane host addressing. `{domain}` and `X-Access-Token` are the `domain` and `sandbox_access_token` from the same responses — no extra endpoint is needed to obtain them. This split is not historical baggage but the nature of a data plane: file transfer is long-lived, high-volume, and addressed per sandbox, differing from the control plane's short requests in both capacity model and failure domain, so they do not share an entry point. **One unified abstraction.** The same contract is served by two backend kinds (an in-sandbox file service / a data-center data-plane proxy); clients need not — and cannot — tell them apart. The cost is that the contract is their intersection: `Content-Length` on download and resumable download (`Range`/`206`) are **optional capabilities** that vary by data center, and clients must work correctly when they are absent. See the per-operation notes below. paths: /templates/{template_id}/builds: parameters: - $ref: '#/components/parameters/TemplateBuildTemplateIdPathParam' get: tags: - SandboxTemplate summary: List Template Builds description: Lists immutable Build attempts in reverse chronological order. operationId: listTemplateBuilds parameters: - $ref: '#/components/parameters/PageParam' - $ref: '#/components/parameters/PageSizeParam' responses: '200': description: Builds were returned successfully. content: application/json: schema: $ref: '#/components/schemas/BuildPage' '400': $ref: '#/components/responses/BadRequest' '401': $ref: '#/components/responses/Unauthorized' '403': $ref: '#/components/responses/Forbidden' '404': $ref: '#/components/responses/NotFound' '500': $ref: '#/components/responses/InternalError' components: parameters: TemplateBuildTemplateIdPathParam: name: template_id in: path required: true description: Owning GMI Template identifier. schema: type: string format: uuid PageParam: name: page in: query description: Page number, starting from 1; default 1 schema: type: integer minimum: 1 default: 1 PageSizeParam: name: page_size in: query description: Items per page; default 20, maximum 100 schema: type: integer minimum: 1 maximum: 100 default: 20 schemas: BuildPage: type: object additionalProperties: false required: - request_id - total - page - page_size - items properties: request_id: type: string description: HTTP request correlation identifier. total: type: integer format: int64 minimum: 0 description: Total number of Build attempts for the Template. page: type: integer minimum: 1 description: One-based page number. page_size: type: integer minimum: 1 maximum: 100 description: Requested number of Builds per page. items: type: array items: $ref: '#/components/schemas/TemplateBuild' TemplateBuild: type: object additionalProperties: false required: - id - template_id - trigger - status - artifact_state - source_type - resources - created_at properties: id: type: string description: Immutable GMI Build identifier. template_id: type: string description: Owning GMI Template identifier. trigger: $ref: '#/components/schemas/BuildTrigger' status: $ref: '#/components/schemas/BuildStatus' artifact_state: $ref: '#/components/schemas/ArtifactState' source_type: $ref: '#/components/schemas/BuildSourceType' source: type: object description: Sanitized Build lineage. Provider IDs and credentials are omitted. additionalProperties: true resources: $ref: '#/components/schemas/EffectiveResources' product: type: string nullable: true description: Catalog SKU when created from a preset; null for custom resources. product_name: allOf: - $ref: '#/components/schemas/ResourceSpecName' nullable: true description: Preset label when created from a preset; null for custom resources. commands: type: array description: Build command strings. Empty for snapshot Builds. items: type: string start_cmd: type: string nullable: true description: Persisted launch command for this Build. failure: $ref: '#/components/schemas/BuildFailure' created_at: type: string format: date-time description: Build creation time in UTC. started_at: type: string format: date-time nullable: true description: Build start time in UTC. finished_at: type: string format: date-time nullable: true description: Build terminal time in UTC. ErrorResponse: type: object required: - code - message - request_id properties: code: type: string description: Machine-readable business error code example: Resource.QuotaExceeded message: type: string description: Human-readable error description example: The resource quota is not enough. request_id: type: string description: Tracing ID example: req-xxxx details: type: object description: Field-level error details additionalProperties: true BuildTrigger: type: string enum: - initial_build - rebuild - snapshot description: Operation that created the Build. BuildStatus: type: string enum: - waiting - building - ready - error description: Normalized Template Build status. ArtifactState: type: string enum: - absent - unknown - available - reclaimed - retained description: | Provider artifact availability independent of Build status. `retained` means the Build was logically deleted but the artifact was left in place; `reclaimed` means the artifact was hard-deleted. BuildSourceType: type: string enum: - image - template - sandbox description: Provider-neutral Build lineage type. EffectiveResources: type: object additionalProperties: false required: - cpu_count - memory_mb - disk_size_mb - architecture properties: cpu_count: type: integer minimum: 1 description: | Effective integer vCPU count. Operator-managed presets do not use the custom-resource maximum. memory_mb: type: integer minimum: 1 description: | Effective memory size in MiB. Operator-managed presets are not constrained by custom-resource limits. disk_size_mb: type: integer minimum: 1 description: | Effective boot disk size in MiB. Operator-managed presets are not constrained by custom-resource limits. architecture: type: string enum: - x86_64 - arm64 description: | CPU architecture resolved for the selected IDC. ResourceSpecName: type: string enum: - X_SMALL - SMALL - MEDIUM - LARGE - X_LARGE description: Stable human-readable Sandbox resource specification label. BuildFailure: type: object nullable: true additionalProperties: false required: - code - message properties: code: type: string description: Stable GMI failure code. example: TemplateBuild.OutOfMemory message: type: string description: Sanitized failure message. responses: BadRequest: description: Invalid request parameters content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: code: Parameter.Invalid message: Invalid request parameters request_id: req-xxxx Unauthorized: description: Unauthenticated content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: code: Auth.Unauthorized message: Authentication required request_id: req-xxxx Forbidden: description: The caller cannot access the organization-scoped resource. content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' NotFound: description: Resource not found content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: code: Resource.NotFound message: The requested resource was not found request_id: req-xxxx InternalError: description: Internal server error content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: code: Internal.ServerError message: An internal server error occurred request_id: req-xxxx securitySchemes: bearerAuth: type: http scheme: bearer description: > API Key or Access Token, always sent as `Authorization: Bearer `. The gateway recognizes the token type automatically. ````