Endpoints Reference - WriterzRoom

The WriterzRoom API exposes generation, status tracking, content management, templates, style profiles, verticals, governance, artifacts, and corpus endpoints through one base URL.

https://api.writerzroom.com/api

Protected account and generation endpoints require Bearer token authentication. Catalog reads for templates, style profiles, and verticals are public, but integrations should still send the key consistently when moving between catalog and protected routes.

Authorization: Bearer YOUR_API_KEY

Each endpoint below also has a generated reference page, with its parameters, request body schema, and an interactive request builder, under API → Endpoint Reference. Those pages are generated from the API server, so they always match what it accepts.

Endpoint Overview

Generation

Submit jobs and poll completion status.

Content

Retrieve, save, and manage generated drafts.

Configuration

Fetch templates, styles, and vertical settings.

Usage

Review account statistics and activity.

Base URL

https://api.writerzroom.com/api

Core Endpoints

Method	Endpoint	Description
`POST`	`/generate`	Submit a generation request
`GET`	`/generate/status/{request_id}`	Poll generation status and retrieve completed content
`GET`	`/generate/stream/{request_id}`	Stream a generation’s progress as server-sent events, resumable with `Last-Event-ID`
`GET`	`/generate/{request_id}/history`	List the checkpoints a generation passed through
`GET`	`/content`	List saved content for the authenticated user
`GET`	`/content/{contentID}`	Retrieve a specific content item
`POST`	`/content/save`	Save or update a content item
`POST`	`/content/regenerate-section`	Rewrite one section of existing content, optionally under an instruction (1 credit)
`GET`	`/templates`	List available templates
`GET`	`/templates/{slug}`	Retrieve a specific template by slug
`GET`	`/style-profiles`	List available style profiles
`GET`	`/style-profiles/{slug}`	Retrieve a specific style profile by slug
`GET`	`/verticals`	List available verticals
`GET`	`/verticals/{vertical_id}`	Retrieve vertical configuration details
`GET`	`/verticals/{vertical_id}/templates`	List the templates a vertical allows and blocks, and its default
`GET`	`/verticals/{vertical_id}/styles`	List the style profiles a vertical allows, recommends, and blocks
`GET`	`/verticals/{vertical_id}/practice-areas`	List the practice-area overlays a request’s `practice_area` parameter can select
`GET`	`/dashboard/stats`	Retrieve usage statistics and generation counts

Streaming a Generation

GET /generate/stream/{request_id} returns text/event-stream. Each event has an id, an event name, and a JSON data payload. If the connection drops, reconnect with the last id you received in the Last-Event-ID header and the stream resumes after it.

Event	Meaning
`stage`	The run moved to another pipeline stage; carries the stage and progress
`agent_event`	A stage recorded activity, such as research found sources
`token`	Draft text as the writer produces it; a preview, not the final content
`preview_reset`	Discard the preview so far; a later stage is rewriting it
`interrupt`	The run paused with the editor’s suggestions for a person to accept or reject
`status`	The run’s state, sent when the stream hands you over to status polling
`complete`	The run finished; fetch the content with `GET /generate/status/{request_id}`
`error`	The run failed; carries the stage and a public reason

A stream served after a restart, or by a different server than the one running the generation, has no token previews: it sends stage events and the final complete, error, or interrupt. If it sends status with "fallback": "poll", continue with GET /generate/status/{request_id}. GET /generate/{request_id}/history returns 501 when the deployment does not keep checkpoints.

Governance Endpoints

Approval routing, post-publication validity monitoring, and replay. See Content Governance for what these records contain and how they are produced.

Method	Endpoint	Description
`GET`	`/approvals`	List the caller’s review queue
`GET`	`/approvals/{request_id}`	Retrieve every review on one asset, with rollup state
`POST`	`/approvals/{request_id}/route`	Derive and persist the reviews an asset requires
`GET`	`/approvals/{approval_id}/asset`	Inspect the content, passport, quality report, and staged review state
`PATCH`	`/approvals/{approval_id}/assign`	Assign or clear a reviewer with the required qualification
`POST`	`/approvals/{approval_id}/decide`	Record a review decision
`GET`	`/validity`	List assets whose sources may no longer hold up
`GET`	`/validity/{request_id}`	Retrieve one asset’s validity record
`POST`	`/validity/{request_id}/scan`	Re-check one asset’s sources now
`POST`	`/validity/{request_id}/acknowledge`	Accept current findings without erasing them
`GET`	`/replay/{request_id}/reproducibility`	Report whether an asset can be reproduced, and how faithfully
`GET`	`/replay/{request_id}/request`	Retrieve the payload that would replay an asset
`POST`	`/replay/{request_id}/compare`	Diff a replayed generation against the original

/replay/{request_id}/request returns a payload for you to submit to /generate; it does not run the generation itself. A replay is metered and rate-limited like any other request.

Artifact Endpoints

Method	Endpoint	Description
`GET`	`/artifacts/types`	Discover artifact types, verifiers, limitations, billing mode, and repair limit
`POST`	`/artifacts/generate`	Generate a strict-schema artifact, verify it, and repair at most once
`POST`	`/artifacts/verify`	Run configured executable verifiers over existing content

See Verified Artifacts API for response semantics and non-claims.

Corpus Endpoints

Method	Endpoint	Description
`POST`	`/corpus/documents`	Upload a tenant-private reference document
`GET`	`/corpus/documents`	List the caller’s private documents
`GET`	`/corpus/documents/stats`	Retrieve private document and chunk counts
`DELETE`	`/corpus/documents/{document_id}`	Delete the caller’s private document chunks
`POST`	`/corpus/search`	Search curated material plus only the caller’s private documents

Shared ingestion and deletion routes are administrator-only. The separate Corpus MCP is read-only and curated-only.

Request Example

curl -X GET https://api.writerzroom.com/api/templates \
  -H "Authorization: Bearer YOUR_API_KEY"

Error Response Example

{
  "detail": "Missing or invalid API key"
}

Integration Flow

Authenticate requests

Add the Authorization: Bearer YOUR_API_KEY header to every protected request.

Fetch valid configuration

Use /templates, /style-profiles, and /verticals to retrieve valid identifiers before submitting generation jobs.

Submit generation

Call POST /generate with the selected template, style profile, vertical, generation mode, and structured input.

Poll for completion

Use /generate/status/{request_id} to retrieve processing updates and completed content.

Save or manage content

Use content endpoints to retrieve, save, update, or regenerate content sections.

Rate Limits

Limit	Value
`POST /generate`	10 submissions per minute per authenticated user
`GET /generate/status/{request_id}`	60 polls per minute per authenticated user
Rate limit response	`429`
Recommended retry strategy	Exponential backoff
Other endpoints	No single universal limit; endpoint-specific protections may apply

Error Codes

Status	Meaning
`400`	Bad request, missing fields, or invalid payload
`401`	Unauthorized, missing or invalid API key
`402`	Insufficient credits or inactive subscription
`403`	Forbidden, plan does not include API access
`404`	Resource not found
`409`	A `request_id` reused for a different generation, or a resume already in progress
`422`	Invalid input, including a malformed `request_id`
`429`	Rate limit exceeded
`500`	Internal server error
`503`	Temporarily unable to start the request; retry it unchanged

Safe Retries

Send your own request_id with POST /generate so a retry cannot start, or charge for, a second generation. Use 8 to 128 letters, digits, hyphens or underscores; a UUID works. The ID must be unique within your account.

Resubmitting the same request_id returns 202 with the original generation’s current status and "duplicate": true. Nothing new is started or charged.
Reusing a request_id for a different template returns 409.
Omitting request_id still works; the API assigns one, but a retried request then counts as a new generation.

Exponential Backoff Example

async function requestWithBackoff(fn, maxRetries = 3) {
  for (let attempt = 0; attempt < maxRetries; attempt++) {
    try {
      return await fn();
    } catch (error) {
      const isRateLimited = error.status === 429;

      if (!isRateLimited || attempt === maxRetries - 1) {
        throw error;
      }

      const delayMs = Math.pow(2, attempt) * 1000;
      await new Promise((resolve) => setTimeout(resolve, delayMs));
    }
  }
}

Operational Guidance

Which base URL should integrations use?

Use https://api.writerzroom.com/api for public API integrations.

Should identifiers be hardcoded?

Avoid hardcoding identifiers for large integrations. Fetch templates, style profiles, and verticals programmatically when possible.

How should rate limits be handled?

Treat 429 as a retryable response and apply exponential backoff. Do not retry immediately in tight loops.

Is it safe to retry a generation request that timed out?

Yes, if the original request carried a request_id: resend it with the same ID and you get the original generation back. See Safe Retries.

What information should be included in support requests?

Include the endpoint, timestamp, request ID if available, response status, and sanitized error body.

Do not send API keys in query parameters. Always use the Authorization header.

API Authentication

Learn how to authenticate API requests.

Check Generation Status

Poll submitted generation jobs until completion.

Dataset analysis inputs

These endpoints require authenticated source-use permission. Inspection and validation share a limit of twelve requests per minute per user. Responses are private and are not cached.

Method	Endpoint	Description
`GET`	`/sources/{version_id}/analysis-schema`	Inspect an exact CSV, JSON, or XLSX version; select a worksheet and header row with query parameters.
`POST`	`/sources/analysis/validate`	Validate a template-enabled analysis request against complete authorized originals before generation.

See Dataset analysis for supported calculations and verification limits.

​Endpoint Overview

Generation

Content

Configuration

Usage

​Base URL

​Core Endpoints

​Streaming a Generation

​Governance Endpoints

​Artifact Endpoints

​Corpus Endpoints

​Request Example

​Error Response Example

​Integration Flow

​Rate Limits

​Error Codes

​Safe Retries

​Exponential Backoff Example

​Operational Guidance

API Authentication

Check Generation Status

​Dataset analysis inputs

Endpoint Overview

Base URL

Core Endpoints

Streaming a Generation

Governance Endpoints

Artifact Endpoints

Corpus Endpoints

Request Example

Error Response Example

Integration Flow

Rate Limits

Error Codes

Safe Retries

Exponential Backoff Example

Operational Guidance

Dataset analysis inputs