---
title: "Export account data"
method: GET
path: "/v1/account/export"
tags: ["Account"]
---

# Export account data

`GET /v1/account/export`

## Request

- Base URL: `https://api.postlake.dev`
- URL: `https://api.postlake.dev/v1/account/export`
- Auth: `bearerAuth` (a scheme the document does not define)

GDPR data export. All data for the account (no secrets).

## Response `200`

Success

- object
  - `account` object, required
    - `id` string, required
    - `email` string, required
  - `profiles` Profile[], required
    - `id` string, required
    - `username` string, required: The identifier agents put in `profile:`. Slugified.
    - `name` string: What the person typed, for display. Absent on profiles created before this existed.
    - `createdAt` string, date-time, required
  - `socialAccounts` SocialAccount[], required
    - `id` string, required
    - `platform` 'bluesky' | 'threads' | 'x' | 'linkedin' | 'instagram' | 'tiktok' | 'facebook' | 'youtube' | 'pinterest', required
    - `handle` string, required
    - `profileId` string, nullable, required
    - `status` string, required
    - `avatarUrl` string, nullable
    - `connectVariant` string: Which door this account was connected through, for networks that have more than one. Instagram is the case: `instagram` means Instagram Login, `facebook` means it was connected through a Facebook Page. The two reach genuinely different feature sets, so this decides what the account can do. Absent on networks with a single way in.
    - `accountType` 'BUSINESS' | 'MEDIA_CREATOR' | 'PERSONAL', nullable: Instagram account type when known. Stories support both login routes, subject to Meta account eligibility. Facebook Login Stories require a Business account. Null means the type is unknown; Meta enforces eligibility during publishing.
    - `discovers` object, required: What this ACCOUNT can search, as opposed to GET /v1/platforms/{platform}, which can only speak for the network as a whole. An Instagram connected by Instagram Login cannot search creators or hashtags however the platform-level flags read, so check here before calling a discovery endpoint rather than learning it from an error.
      - `posts` boolean, required
      - `profiles` boolean, required
      - `profilePosts` boolean, required
      - `places` boolean, required
      - `creators` boolean, required: True only when the connection can reach its creator-discovery product. A true still leaves the network's own onboarding to be done: Meta refuses the search with `creator_marketplace_not_onboarded` until the brand has accepted the relevant terms.
    - `requires` string, nullable: What this network needs before it will publish, in one sentence (media requirements, title and caption limits). Read it before composing: TikTok, Instagram, YouTube and Pinterest all refuse text-only posts. Null when the network accepts anything.
  - `posts` Post[], required
    - `id` string, required
    - `sourceDraftId` string: Immutable original draft ID. New approvals preserve this as id; legacy audited approvals may resolve an older ID to a different canonical id.
    - `snapshot` string: Computed full-state approval fence returned by GET. Includes media, options and metadata.
    - `state` 'draft' | 'awaiting_approval' | 'rejected' | 'queued' | 'scheduled' | 'processing' | 'partial' | 'published' | 'failed', required
    - `createdAt` string, date-time, required
    - `clientId` string: Stable machine principal that created the post. Absent for human-created dashboard posts.
    - `surface` 'dashboard' | 'api' | 'agent': Where the post was created, independent of the principal identity.
    - `approval` object: Present for approval-gated agent work, including the owner's final decision when one has been made.
      - `requestedAt` string, date-time, required
      - `decision` 'approved' | 'rejected'
      - `decidedAt` string, date-time
      - `reason` string
    - `scheduledAt` string, date-time, nullable, required
    - `timezone` string: IANA timezone the caller scheduled in, when they named one.
    - `scheduledAtLocal` string: Wall-clock time in timezone, computed on read. Not stored.
    - `text` string, required
    - `campaign` string
    - `assetMetadata` object
    - `media` string[], required
    - `warnings` string[]: Advisory notes from create or edit, for example PNG converted to JPEG for TikTok.
    - `platformOptions` object, nullable, required
      - `pinterest` object
        - `boardId` string
        - `link` string
        - `altText` string
    - `targets` PostTarget[], required
      - `removal` object: Present after accepted network removal on a partial result. Retry skips accepted destinations; verified separately confirms absence.
        - `accepted` true, required
        - `verified` boolean, required
        - `at` string, date-time, required
      - `account` string, required
      - `platform` 'bluesky' | 'threads' | 'x' | 'linkedin' | 'instagram' | 'tiktok' | 'facebook' | 'youtube' | 'pinterest', required
      - `state` 'queued' | 'scheduled' | 'processing' | 'published' | 'failed', required
      - `profile` string: Username of the profile this channel belongs to. Resolved when the post is read, so it follows the channel if it moves. Absent once the channel is disconnected.
      - `channelName` string: The channel's handle at the time of reading, not the one frozen at publish. Absent once the channel is disconnected.
      - `connection` object: The native account or destination identity captured when this target was scheduled. It is used only to repair a disconnected channel safely.
        - `platformAccountId` string
        - `destinationId` string
        - `destinationName` string
        - `handle` string
      - `remoteId` string, nullable, required
      - `url` string, nullable, required
      - `publishedAt` string, date-time, nullable, required
      - `permalinkPending` boolean: True when publication succeeded but an exact public permalink is still being resolved.
      - `error` NormalisedError, required: `type` is the coarse category you route on and is deliberately small, so it cannot say WHICH input was wrong. `code`, `fix`, `docs` and `param` are what turn a rejection into something a caller can act on rather than guess at. They are optional only in the sense that not every error has all four.
        - `type` string, required: Coarse category to route on, e.g. rate_limited, auth_expired, platform_rejected.
        - `message` string, required
        - `platform` 'bluesky' | 'threads' | 'x' | 'linkedin' | 'instagram' | 'tiktok' | 'facebook' | 'youtube' | 'pinterest'
        - `retryable` boolean, required
        - `code` string: Stable, granular handle to branch on, e.g. text_too_long. Also the anchor in the error reference.
        - `fix` string: The next action, in plain words. Safe to show a person as-is.
        - `docs` string, uri: Where the rule is written down, so it is learned once rather than rediscovered.
        - `param` string: The request field at fault, so a form can mark the right input.
  - `auditLog` AuditEvent[], required
    - `id` string, required
    - `action` string, required
    - `ip` string, required
    - `userAgent` string, required
    - `createdAt` string, date-time, required
  - `exportedAt` string, date-time, required

## Other responses

- `401`: Unauthorized

---

[API](https://skmtc.dev/postlake/apis/postlake-api.md) · [All operations](https://skmtc.dev/postlake/apis/postlake-api/llms.txt) · [OpenAPI document](https://skmtc.dev/postlake/apis/postlake-api/revisions/f81060e31f69?raw)