BUILD stamp logo
FORM CIBB-MUSE[Connector v1]

Muse connector

The connector turns a hardware idea into a private manufacturing-feasibility assessment. It uses the same assessment engine as canitbebuilt.ai. Version 1 does not accept payments, publish ideas, or generate a public Factory Floor page.

Base URL

https://www.canitbebuilt.ai/api/connectors/muse/v1

  • GET /health
  • POST /assessments
  • GET /assessments/{assessmentId}
  • GET /openapi.json

Authentication

Send Authorization: Bearer with a key created while signed in at API keys. The full key is shown once and looks like cibb_live_.... Keys can be revoked. A revoked key is rejected. The connector does not accept a browser CAPTCHA token.

Scopes on a new key: assessment:create and assessment:read. One key cannot read another owner's assessment.

Example request

curl -s -X POST https://www.canitbebuilt.ai/api/connectors/muse/v1/assessments \
  -H "Authorization: Bearer cibb_live_..." \
  -H "Content-Type: application/json" \
  -d '{   "productIdea": "A screenless sleep ring with a seven-day battery, retail under 300 USD.",   "targetRetail": { "currency": "USD", "minimum": 0, "maximum": 300 },   "clientRequestId": "muse-example-001" }'

Required: productIdea (10 to 8,000 characters). Optional: audience, targetRetail (currency, minimum, and optional maximum), targetMarket, additionalDetails, sketchUrl (HTTPS PNG, JPEG, or WebP), and clientRequestId.

Example response

A finished assessment returns HTTP 201. The body includes the verdict (BUILDABLE, BUILDABLE_WITH_CHANGES, HEAVY_LIFT, or NOT_VIABLE), a short explanation, assumptions, subsystem risks, BOM cost strings at 100, 1,000, and 10,000 units with an explicit currency, tooling, DFM, certification, and tooling gates, a prototype sequence, unit-economics text, warnings, and the AI disclaimer. A user API key also receives continueUrl, the absolute private report URL for that account. The operator review key omits continueUrl.

If the same clientRequestId is still processing, the POST returns HTTP 202 with assessmentId, status, pollAfterSeconds, and statusUrl. GET that URL for HTTP 200 with status queued, processing, completed, or failed.

This is a system generated assessment issued on the basis of the information furnished above. It is produced by an automated model and may be incomplete, inaccurate, or fabricated. It is not a substitute for a DFM review by a manufacturer, nor for independent engineering, legal, or manufacturing due diligence.

Rate limits

Current configuration: 30 requests per minute and 500 requests per day, counted per API key and per network signal. Sketch downloads are limited to about 5 MB. HTTP 429 responses include a Retry-After header.

Data handling

Connector assessments are private. They are not placed on the Factory Floor unless you later publish them while signed in. Repeating a clientRequestId with the same payload returns the original assessment. A different payload with that id returns HTTP 409. Details are in the privacy policy.

Error codes

  • invalid_request
  • authentication_required
  • invalid_api_key
  • insufficient_scope
  • rate_limited
  • idempotency_conflict
  • assessment_not_found
  • sketch_fetch_failed
  • assessment_failed
  • service_unavailable

Every error uses { "error": { "code", "message", "retryable", "requestId" } }. Responses do not include stack traces, prompts, or provider payloads.

Support

inspector@canitbebuilt.ai. See also the terms.