---
title: "get_status for a quote, with a message to the operator"
method: POST
path: "/quotes/{quote_id}"
---

# get_status for a quote, with a message to the operator

`POST /quotes/{quote_id}`

## Request

- Base URL: `https://api.handsforagents.com/v1`
- URL: `https://api.handsforagents.com/v1/quotes/{quote_id}`
- Auth: HTTP bearer

This is get_status with the optional message, as a POST so that the text is in the body, not in the URL (which servers log). Without a message (empty body, or {"message": null}) it is a plain status call, like the GET. With a message it is written to the thread and e-mailed to the operator, and the answer is the same as the GET's, including your message (from: agent) and message_sent: true. Read the operator's message first, then reply: while the operator has written something you have not read yet, nothing is stored or e-mailed; the answer is the GET's, with that message in messages, message_sent: false and a next_action that says so, and you send your answer with the next call (HTTP 200 either way, so check message_sent). Plain text, 1-2000 characters, never interpreted as HTML; an empty or longer message is a 400. Limits: 10 messages per hour per quote (its quote_id and task_id share it) and 30 per hour per client address (429). The same text again within 10 minutes, while it is still the newest message of the thread, is not stored or e-mailed a second time (the answer then has duplicate_ignored: true). It needs the same access_token as the GET; without one, or with another quote's, nothing is stored or sent (401).

## Path parameters

- `quote_id` string, required

## Request body

- StatusMessage: Optional body of POST /quotes/{quote_id} and POST /tasks/{task_id}. An empty body is a plain status call.
  - `message` string, nullable: Your reply to the operator: plain text, 1-2000 characters, never HTML. Null or absent: a plain status call.

## Response `200`

The same answer as the GET, now including your message; if the operator had written something you had not read, message_sent is false, nothing was stored and the answer shows that message

- Quote
  - `quote_id` string, required
  - `status` 'received' | 'quoted' | 'refused' | 'expired', required
  - `refusal_reason` string: Present when status is refused. Refused requests are not invoiced.
  - `refusal_code` 'credentials' | 'impersonation' | 'surveillance' | 'social-manipulation' | 'verification-bypass' | 'referral-fraud' | 'weapons-regulated' | 'counterfeit' | 'reshipping-fraud' | 'illegal' | 'capability-limit' | 'sensitive-unverified' | 'other': Present when status is refused: the category from the task policy, or capability-limit / sensitive-unverified / other.
  - `task_id` string: Present once create_task was called for this quote.
  - `price` object
    - `amount` number, required: Fixed total. No VAT is added; the provider is not VAT-registered.
    - `currency` 'EUR' | 'USD', required
  - `payment_schedule` object: Up to 150 EUR: full_prepayment or card_hold. Above: deposit_and_balance.
    - `mode` 'full_prepayment' | 'card_hold' | 'deposit_and_balance', required
    - `deposit` object
      - `amount` number, required: Fixed total. No VAT is added; the provider is not VAT-registered.
      - `currency` 'EUR' | 'USD', required
    - `balance` object
      - `amount` number, required: Fixed total. No VAT is added; the provider is not VAT-registered.
      - `currency` 'EUR' | 'USD', required
  - `scope` string
  - `deliverables` string
  - `evidence` string[]: What the evidence package will contain.
  - `subcontracted` string: Processes done by subcontractors, if any.
  - `delivery_date` string, date
  - `inbound_address` string: Where to send parts or parcels, if the task needs them.
  - `operator_verification_required` boolean: True for the sensitive task categories.
  - `valid_until` string, date
  - `terms_version` string
  - `next_action` string: One or two sentences in English: what the agent or the client should do now.
  - `messages` object[]: The thread with the operator, newest first: the operator's notes and your own replies (from: agent over REST; the MCP tools show your own replies as from: you). An operator message you have not been shown yet is always included, also in the answer to your own reply (a message of yours sent while it is unread is not stored: message_sent is false). Read the operator's message first, then reply: call get_status with message (POST with a JSON body over REST); e-mailing the tasks address with the quote_id in the subject also works.
    - `at` string, date-time, required
    - `from` 'operator' | 'agent', required
    - `text` string, required: Plain text, at most 2000 characters. Never HTML.
  - `duplicate_ignored` boolean: Present and true only in the answer to a POST whose message was ignored because the same text was already the newest message of the thread (sent within the last 10 minutes).
  - `message_sent` boolean: Present only in the answer to a POST with a message: true, the message was stored and e-mailed to the operator; false, it was NOT stored because the operator had written something you had not read yet (that message is in messages, now marked as read; next_action says so): send your answer again. Absent when the same text was ignored (duplicate_ignored) and in every plain status call.
  - `test_mode` boolean: Present and true only for a test quote: an order made from it is a test order.

## Other responses

- `400`: Error
- `401`: Error
- `404`: Error
- `429`: Error

---

[API](https://skmtc.dev/handsforagents/apis/hands-for-agents-api.md) · [All operations](https://skmtc.dev/handsforagents/apis/hands-for-agents-api/llms.txt) · [OpenAPI document](https://skmtc.dev/handsforagents/apis/hands-for-agents-api/revisions/8349b6792c56?raw)