Developers › API reference › AI
POST /api/v1/commands/{screen}
Plan a natural-language command and preview it (screen: portfolio, blotter, monitor, wfrules); nothing runs.
- Scope
ai:use - Idempotency-Key
- operationId
aiPlan
The token needs the scope ai:use, and the person or service account behind it needs the matching privileges and account permissions, exactly as on the screens. See Scopes.
Idempotency-Key. Send a unique Idempotency-Key (up to 255 characters, e.g. a UUID) with every POST. Retrying with the same key and body replays the first answer for 24 hours, so a timeout never creates a second order; the same key with a different body is refused (idempotency_key_reused).
Request
POST /api/v1/commands/{screen}
Path parameters
| Name | Description | Example |
|---|---|---|
screen | portfolio, blotter, monitor or wfrules. | blotter |
Headers
| Header | Required | Value |
|---|---|---|
Authorization | yes | Bearer <token>: an OAuth access token (vat_…) or a personal access token (vpt_…). Get a token |
Idempotency-Key | yes | A unique string up to 255 characters (a UUID). Idempotency |
Content-Type | yes | application/json |
Body: CommandIn
| Field | Type | Required | Description |
|---|---|---|---|
text | string | yes | The command in plain English. |
answer_to | any | no | The id of the command whose clarifying question you are answering. |
Decimals (money, prices, quantities, percentages) are sent as strings, ids as strings, times as ISO 8601 UTC.
Examples
Keep the token in an environment variable (VESTORIS_TOKEN), never in the script.
curl
curl -X POST "https://vestoris.com/api/v1/commands/blotter" \
-H "Authorization: Bearer $VESTORIS_TOKEN" \
-H 'Idempotency-Key: 7f9c2b1e-4d0a-4a7e-9a51-3c2f0b8d6e11' \
-H "Content-Type: application/json" \
-d '{"text":"Cancel all my unfilled AAPL buys"}'
Python (requests)
import os, requests
r = requests.post(
"https://vestoris.com/api/v1/commands/blotter",
headers={
"Authorization": f"Bearer {os.environ['VESTORIS_TOKEN']}",
"Idempotency-Key": "7f9c2b1e-4d0a-4a7e-9a51-3c2f0b8d6e11",
},
json={
"text": "Cancel all my unfilled AAPL buys"
},
timeout=30,
)
r.raise_for_status() # errors are application/problem+json with a stable "code"
print(r.json())
JavaScript (fetch)
const res = await fetch("https://vestoris.com/api/v1/commands/blotter", {
method: "POST",
headers: {
Authorization: `Bearer ${process.env.VESTORIS_TOKEN}`,
"Idempotency-Key": "7f9c2b1e-4d0a-4a7e-9a51-3c2f0b8d6e11",
"Content-Type": "application/json",
},
body: JSON.stringify({
"text": "Cancel all my unfilled AAPL buys"
}),
});
const data = await res.json();
if (!res.ok) throw new Error(`${data.code}: ${data.detail}`); // problem+json
console.log(data);
Response
201 Created, application/json: Command. Rate-limit headers: RateLimit-Limit, RateLimit-Remaining, RateLimit-Reset.
Response fields
| Field | Type | Description |
|---|---|---|
id | id (string) | The record's id (a string of digits). |
screen | string | The screen the command is for: portfolio, blotter, monitor or wfrules. |
text | string | The command in the user's words. |
status | string | planned, question, previewed, executed, cancelled, stale or failed. |
planned_by_ai | boolean | Planned by ai. |
question | string or null | A clarifying question, when something material is ambiguous. |
message | string | What Vestoris says about the plan. |
plan | object | The typed actions the command was planned as. |
preview | object or null | What would happen: each item before and after, what is left out and why. |
item_count | integer | How many items the plan touches. |
confirm_token | string or null | Send this in X-Vestoris-Confirm to confirm exactly what was previewed. |
typed_confirmation | string or null | Text to send as "typed" for large or destructive plans. |
results | list | Each item's result. |
created_at | date-time or null | When it was created. |
previewed_at | date-time or null | When the preview was made. |
executed_at | date-time or null | When the plan was executed. |
Example response
HTTP/1.1 201 Created
Content-Type: application/json
{
"id": "55",
"screen": "blotter",
"text": "Cancel all my unfilled AAPL buys",
"status": "previewed",
"planned_by_ai": true,
"question": null,
"message": "3 orders would be cancelled; 1 is left out (already filled).",
"plan": {
"actions": [
{
"type": "cancel_orders",
"filter": {
"side": "buy",
"symbols": [
"AAPL"
],
"mine": true
}
}
]
},
"preview": null,
"item_count": 3,
"confirm_token": "vcf_c41e…",
"typed_confirmation": "CONFIRM 3",
"results": [],
"created_at": "2026-10-27T14:30:00Z",
"previewed_at": null,
"executed_at": null
}
Example values. Your values depend on your firm's data; prices, orders and fills are simulated.
Status codes and errors
Errors are application/problem+json (RFC 9457) with a stable code to program against; the detail says what to fix. See Errors.
| Status | When | Problem codes |
|---|---|---|
| 201 | Created | |
| 200 / 201 | The same Idempotency-Key again: the first answer is replayed (24 hours) | |
| 400 | The request is malformed | invalid_request invalid_json |
| 401 | No token, or it is invalid or expired | invalid_token |
| 403 | The token lacks the scope, or the person or service account behind it lacks the privilege or account permission | insufficient_scope forbidden ip_not_allowed |
| 404 | Not found, or not visible to you (the API never confirms what you may not see) | not_found |
| 409 | Not possible in the current state | invalid_transition conflict idempotency_in_progress |
| 422 | Validation failed (the detail says which field), or blocked | validation_failed idempotency_key_reused |
| 428 | A required header is missing | idempotency_key_required |
| 429 | Rate limited: wait for Retry-After seconds | rate_limited |
{
"type": "https://vestoris.com/developers/errors#validation_failed",
"title": "Validation failed",
"status": 422,
"detail": "Quantity must be greater than zero.",
"code": "validation_failed",
"instance": "/api/v1/commands/blotter"
}
On the screens
The same thing in the app is explained in: