Generate items from a prompt

POST /boards/{boardId}/ai/generate

Updated 8 Oct 20263 min read

POST/boards/{boardId}/ai/generateTry it

Basehttps://api.mockflow.com/v1Scopesai:generateboards:write

Request

curl https://api.mockflow.com/v1/boards/brd_3f9e2a/ai/generate \
  -X POST \
  -H "Authorization: Bearer $MOCKFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "prompt": "Plan the launch of our mobile app. Timeline, kanban and a risk mind map.",
  "outputTypes": [
    "timeline",
    "kanban",
    "mindmap"
  ]
}'

const response = await fetch("https://api.mockflow.com/v1/boards/brd_3f9e2a/ai/generate", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.MOCKFLOW_API_KEY}`,
    "Content-Type": "application/json"
  },
  body: JSON.stringify({
  "prompt": "Plan the launch of our mobile app. Timeline, kanban and a risk mind map.",
  "outputTypes": [
    "timeline",
    "kanban",
    "mindmap"
  ]
})
});

const data = await response.json();

import os, requests

response = requests.post(
    "https://api.mockflow.com/v1/boards/brd_3f9e2a/ai/generate",
    headers={"Authorization": f"Bearer {os.environ['MOCKFLOW_API_KEY']}"},
    json={
  "prompt": "Plan the launch of our mobile app. Timeline, kanban and a risk mind map.",
  "outputTypes": [
    "timeline",
    "kanban",
    "mindmap"
  ]
},
)

response.raise_for_status()
data = response.json()

Response

{
  "dryRun": false,
  "estimate": {
    "items": 0,
    "credits": 0,
    "note": "string"
  },
  "credits": {
    "remaining": 0,
    "total": 0
  },
  "plan": "string",
  "canGenerate": false
}

{
  "id": "job_7hd2",
  "kind": "item.create",
  "status": "queued",
  "progress": 0,
  "message": "Interview 5 users",
  "result": {
    "itemIds": [
      "string"
    ],
    "updated": [
      "string"
    ],
    "deleted": [
      "string"
    ],
    "creditsUsed": 0,
    "boardTitle": "Q4 Launch Planning",
    "skill": "string"
  }
}

{
  "error": {
    "code": "plan_limit",
    "message": "AI credits exhausted. 0 of 2000 remaining this month.",
    "upgradeUrl": "https://mockflow.com/pricing"
  },
  "requestId": "req_01ja"
}

{
  "error": {
    "code": "rate_limited",
    "message": "Too many requests.",
    "docsUrl": "https://mockflow.com/developers/concepts/errors#rate_limited"
  },
  "requestId": "req_45d49816999bac40"
}

Mida turns a prompt into one or more items on the board. With one outputTypes entry the matching generator runs directly; with none, Mida's planner decides what to build, as it does in the editor. skill applies a saved skill's instructions and output type. withImages lets Mida generate pictures inside the result. Plus and Max only; charged in AI credits. dryRun returns an estimate without spending anything.

Authorization

Authorizationstringheaderrequired

Bearer <key>. The key needs ai:generate and boards:write scopes. Create and manage keys in the developer console, or use an OAuth app to act for other people; what each scope unlocks is on the scopes page.

Path parameters

boardIdstringpathrequired

Headers

Idempotency-Keystringheader

A unique key per logical request. Replays within 24 hours return the original response.

Body

application/json

callbackUrlstring

Optional. When the job this request starts finishes or fails, a short body is POSTed here once: { eventId, status, boardId, boardUrl, imageUrl }, where status is succeeded or failed and imageUrl is a picture of what the job drew, or null. Signed with the workspace's callback secret (GET /webhooks/callback-secret). https to a public host. The full job is on GET /jobs/{jobId}.

promptstringrequired

outputTypesenum<string>[]

skillstring

A skill id or slug from GET /skills.

withImagesboolean

modelstring

A text model id from the editor's model picker. Omit for Mida's default.

dryRunboolean

Defaults to false.

Response

200 · application/json

dryRunboolean

estimateobject

Child attributes

itemsinteger

creditsinteger

notestring

creditsobject

Child attributes

remaininginteger

totalinteger

planstring

canGenerateboolean

Errors

402plan_limit

The workspace plan does not allow this, or AI credits are exhausted.

429rate_limited

Too many requests.

Was this page helpful?