> ## Documentation Index > Fetch the complete documentation index at: https://docs.gpuoutlet.ai/llms.txt > Use this file to discover all available pages before exploring further. # Ask about a node > Files an inquiry the team follows up by hand. Requires `radar:write`. `Idempotency-Key` is accepted and echoed but not required here. ## OpenAPI ````yaml https://api.gpuoutlet.ai/v1/openapi.yaml post /radar/nodes/{id}/inquiries openapi: 3.1.0 info: title: GPU Outlet Partner API version: 2.0.0 description: > Programmatic access to the GPU Outlet introduction marketplace: state what capacity your company needs, read the opportunities offered to you, follow the introductions you are party to, and consult Radar. **Authentication.** Every endpoint except this document requires `Authorization: Bearer gpk_…`. Mint a key in your dashboard under Settings → API keys. The key is shown once, at creation, and is not recoverable — we store only its hash. **Company context.** A key belongs to a person; the company and the role the call acts with are that person's company membership, exactly as in the dashboard. A key whose owner is in no company gets `company_required`; a role short of the action gets `forbidden_role`. Nothing in a request body can name a company. **Scopes.** Each endpoint names the scope it needs. A key carries the scopes it was minted with and cannot grow one later; a key without the required scope gets `insufficient_scope` with `required_scope` in the error. **Idempotency.** Every `POST`, `PUT` and `PATCH` under Requests, Opportunities and Introductions **requires** `Idempotency-Key`. Generate one per intent (a UUID is fine) and reuse it verbatim on every retry of that same request: you get the original response back, marked `Idempotency-Replayed: true`, and nothing is written twice. The same key with a different body is `idempotency_key_reused`; a key whose first attempt is still running is `idempotency_key_in_flight` with `Retry-After`. Keys are remembered for 24 hours. Radar writes accept the header and echo it but do not require it. **Rate limits.** Every response carries `RateLimit-Limit`, `RateLimit-Remaining` and `RateLimit-Reset`; a 429 adds `Retry-After`. Writes have a smaller bucket of their own on top. Limits are enforced per API server instance and are approximate — treat `RateLimit-Remaining` as advisory, and back off on 429 rather than predicting it. **Compatibility.** We may add endpoints, add response fields, add optional query parameters, and add values to response enums at any time. Your client **must ignore fields it does not recognise.** Removing or renaming a field, changing a type, or tightening validation requires a new major version. **UI only.** The steps with legal effect stay in the dashboard, behind a second factor a key cannot present: accepting an opportunity and revealing the buyer, existing-relationship claims, deal reports and receipts, deal and success-fee invoices, and signing agreements. This API reads what those steps produced; it does not perform them. **What we do not publish.** Before a reveal, an opportunity carries the buyer's demand and nothing about the buyer — no name, no country, no id. Internal identifiers of people are never on the wire. Do not build on inferring either. servers: - url: https://api.gpuoutlet.ai/v1 security: - bearerAuth: [] tags: - name: Account description: Facts about the calling key. - name: Radar description: > Curated nodes of reserved capacity — what a signed-in buyer sees on the Radar page, and what a seller has proposed to it. - name: Requests description: > Your company's demand. A request is what the operator matches sellers against; sellers see an anonymised copy of it as an opportunity. - name: Opportunities description: > The seller's inbox — demand matched to your listing, anonymised until you accept it in the dashboard. Declining needs no second factor and is here. - name: Introductions description: > What both parties read after a reveal, and the one move either may make without a person present — "we are ready". paths: /radar/nodes/{id}/inquiries: post: tags: - Radar summary: Ask about a node description: > Files an inquiry the team follows up by hand. Requires `radar:write`. `Idempotency-Key` is accepted and echoed but not required here. operationId: createRadarInquiry parameters: - name: id in: path required: true schema: type: string - $ref: '#/components/parameters/OptionalIdempotencyKey' requestBody: required: true content: application/json: schema: $ref: '#/components/schemas/RadarInquiryInput' responses: '201': description: The inquiry was filed content: application/json: schema: $ref: '#/components/schemas/RadarInquiry' '400': $ref: '#/components/responses/BadRequest' '401': $ref: '#/components/responses/Unauthorized' '403': $ref: '#/components/responses/Forbidden' '404': $ref: '#/components/responses/NotFound' '429': $ref: '#/components/responses/RateLimited' components: parameters: OptionalIdempotencyKey: name: Idempotency-Key in: header required: false schema: type: string minLength: 1 maxLength: 255 pattern: ^[\x20-\x7E]+$ description: Validated and echoed on Radar writes; not stored there. schemas: RadarInquiryInput: type: object additionalProperties: false required: - contact_name - contact_email - desired_duration properties: contact_name: type: string minLength: 1 maxLength: 200 contact_email: type: string format: email maxLength: 320 contact_phone: type: - string - 'null' maxLength: 60 desired_duration: type: string minLength: 1 maxLength: 200 description: Free text comments: type: - string - 'null' maxLength: 2000 RadarInquiry: type: object required: - id - status - node_id properties: id: type: string status: type: string enum: - pending - contacted - closed node_id: type: string Error: type: object required: - error properties: error: type: object required: - type - code - message - doc_url - request_id properties: type: type: string enum: - invalid_request_error - authentication_error - permission_error - rate_limit_error - api_error code: type: string description: Stable machine-readable code message: type: string param: type: string description: The offending parameter, when applicable doc_url: type: string format: uri request_id: type: string description: >- Quote this when contacting support — it locates the exact request required_scope: type: string description: 'On `insufficient_scope`: the scope this endpoint needs' retry_after_seconds: type: integer description: >- On a rate-limit error or `idempotency_key_in_flight`: seconds to wait current: type: string description: > On `request_not_editable`, `opportunity_not_pending` and `introduction_not_active`: the status the resource is in. problems: type: array items: type: string description: >- On `request_invalid`: every field at fault (`gpu_model`, `term`, …) responses: BadRequest: description: >- `invalid_request` — a parameter was invalid (`param` names it) — or `idempotency_key_required` content: application/json: schema: $ref: '#/components/schemas/Error' Unauthorized: description: Missing, malformed, unknown, revoked or expired key content: application/json: schema: $ref: '#/components/schemas/Error' Forbidden: description: > The key is valid but not allowed to do this: `insufficient_scope` (with `required_scope`), `company_required`, `forbidden_role`, `api_keys_disabled`, or `banned`. content: application/json: schema: $ref: '#/components/schemas/Error' NotFound: description: '`not_found` — no such resource, or it is not visible to you' content: application/json: schema: $ref: '#/components/schemas/Error' RateLimited: description: Rate limit or daily quota spent headers: Retry-After: schema: type: integer description: Seconds to wait before retrying content: application/json: schema: $ref: '#/components/schemas/Error' securitySchemes: bearerAuth: type: http scheme: bearer description: 'An API key from your dashboard, sent as `Authorization: Bearer gpk_…`.' ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.