Browse documentation

Read evidence, render pages and publish reviewed articles

Use scoped API keys for website evidence, managed rendering, full article generation and reviewed publishing workflows.

On this page

Overview

Use Prerender Buddy's Developer API from backend services, CI jobs and trusted tools. Read scoped website and AI Visibility evidence, use managed rendering and diagnostics, and prepare, generate, review, deliver or schedule articles through the documented workflows. Existing account allowances and explicit action confirmations apply.

Read the Markdown reference · Open the API contract

Access follows your workspace and key scopes

Developer API access is included in Starter, Growth and Pro. Legacy workspaces retain their effective subscription entitlements. Create a key from Developer API keys in the account avatar menu.

Create a scoped Developer API key in your account and keep it server-side. Each operation requires its documented scope. Keys and site requests remain bounded by the workspace and its effective site allowance.

Start with a read

Use the sites scope to list registered sites available to the workspace:

bash

curl --fail-with-body \
  -H "Authorization: Bearer $PRERENDER_BUDDY_API_KEY" \
  https://api.prerenderbuddy.com/v1/developer/sites

PRERENDER_BUDDY_API_KEY is an example local variable containing your key. The documented response includes sites, plan and requestId. Use an actual returned site UUID when requesting evidence for a selected site.

Read saved evidence

The six site-specific read operations cover:

Overview: setup, monitoring, activity and AI visibility context.
Health: bounded health, incident, discovery-file and readiness evidence.
Crawler activity: recorded activity grouped by platform/page, excluding synthetic checks.
AI visibility: aggregated provider, source-domain and tracked-competitor evidence.
Recommendations: evidence-grounded next actions for the latest completed collection.
Content: bounded calendar and article-draft metadata.

These endpoints read evidence; they do not start collections, publish articles or apply recommendations. Health reads exclude full page HTML; visibility reads exclude full provider-answer bodies; content reads exclude draft bodies. Missing or unavailable evidence is not proof of a negative outcome. The article-draft retrieval endpoint below returns the article text and review hash; its scope is different from the bounded content-status read.

Endpoint and scope reference

Endpoint	Required scope	Purpose
`GET /v1/developer/sites`	`sites`	List registered sites available to the API key workspace.
`GET /v1/developer/usage`	`usage`	Get current plan, quota, and Developer API usage.
`GET /v1/developer/logs`	`usage`	Get recent render/API logs.
`GET /v1/developer/sites/{id}/overview`	`sites`	Get a concise setup, monitoring, activity, and AI visibility overview for one site.
`GET /v1/developer/sites/{id}/health`	`health`	Get bounded website-health, incident, discovery-file, and readiness evidence.
`GET /v1/developer/sites/{id}/crawler-activity`	`activity`	Get recent real crawler activity grouped by platform and page.
`GET /v1/developer/sites/{id}/visibility`	`visibility`	Get AI visibility results by platform, source domain, and tracked competitor.
`GET /v1/developer/sites/{id}/recommendations`	`visibility`	Get evidence-grounded visibility recommendations for the latest completed collection.
`GET /v1/developer/sites/{id}/content`	`content`	List bounded content-calendar and article-draft metadata for one site.
`POST /v1/developer/render`	`render`	Render one registered public URL through the managed Prerender Buddy platform.
`POST /v1/developer/diagnostics`	`diagnostics`	Run crawler visibility diagnostics for one registered public URL.
`POST /v1/developer/cache/purge`	`cache`	Purge a cached render for one registered public URL.
`POST /v1/developer/sites/{id}/verify`	`sites`	Verify whether crawler traffic reaches Prerender Buddy for a registered site.

Render a registered public page

The managed render endpoint can return metadata and crawler-readable HTML. Replace the illustrative URL below with a public URL registered to your workspace:

bash

curl --fail-with-body \
  -X POST https://api.prerenderbuddy.com/v1/developer/render \
  -H "Authorization: Bearer $PRERENDER_BUDDY_API_KEY" \
  -H "Content-Type: application/json" \
  --data '{"url":"https://example.com/pricing","includeHtml":true}'

Fresh API renders use the workspace render allowance; cache hits do not. A fresh rendered diagnostic result may also consume allowance. Set includeHtml to false when your client does not need the HTML in the response.

Inspect cache, usage and failures

Read current usage and recent logs from your workflow. Cache purge is a mutating operation for a registered URL; invalidation and obtaining a new render are separate actions.

Documented API errors use an error object with code and message, plus a requestId. Authentication failures, scope/entitlement failures, site-limit failures and fresh-render quota exhaustion require different actions. Keep the request ID for troubleshooting and remove credentials or sensitive content from logs.

Only public HTTP/HTTPS URLs are accepted for URL operations. Private destinations, local hostnames, unsafe redirects and excessive workloads are blocked or bounded. Exact site setup and verification values belong in your authenticated setup instructions.

Choose the reference for your task

Markdown reference: all paths, request fields, scopes and documented errors.
Live OpenAPI: machine-readable contract for API tooling.
Developer tools: local diagnostics and assistant integrations.
Self-hosting: operate the standalone engine in your own infrastructure.

Reviewed article workflows

Connect n8n, Zapier, Make and Activepieces to the article workflow. Use content for reading, add content:write for proposal preparation and generation, and add content:publish for approval, delivery and scheduling.

Method and endpoint	Required scopes	Purpose
GET `/v1/developer/sites/{siteId}/content/ideas`	`content`	List ideas from retained tracking evidence
POST `/v1/developer/sites/{siteId}/content/proposals`	`content`, `content:write`	Prepare a sourced proposal; retain the request UUID on retries
GET `/v1/developer/sites/{siteId}/content/tasks/{taskId}`	`content`	Read the proposal, task status, allowance and completed draft
POST `/v1/developer/sites/{siteId}/content/tasks/{taskId}/generate`	`content`, `content:write`	Generate a confirmed draft using one shared allowance unit
GET `/v1/developer/sites/{siteId}/content/drafts/{draftId}`	`content`	Read the article and its exact review hash
POST `/v1/developer/sites/{siteId}/content/drafts/{draftId}/approve`	`content`, `content:publish`	Approve the reviewed version
POST `/v1/developer/sites/{siteId}/content/drafts/{draftId}/deliver`	`content`, `content:publish`	Send a CMS draft or explicitly publish
PUT `/v1/developer/sites/{siteId}/content/drafts/{draftId}/schedule`	`content`, `content:publish`	Schedule the approved version or cancel

Generation, approval, live publication and scheduling require explicit confirmations. These actions do not authorize changes to billing, website configuration or monitoring settings. A changed article must be reviewed again. See the live OpenAPI contract for request and response schemas.

See Article Generation for the evidence and editorial workflow, Integrations for destination behavior, and How PB Works for the complete platform. API publishing capability is separate from Buddy and the narrower MCP tool surfaces.