Public API
Read AbuseZero's prices, docs, references and legal documents, contact AbuseZero and start agent onboarding, with no key or sign-in.
The public API answers without a key or a sign-in. It's how an agent reads what it needs before there's an account: prices, the developer docs, the API's references, legal documents and how to reach AbuseZero. It's also where an agent sets up AbuseZero for someone new. Everything comes as JSON, and pages come as Markdown.
How the public API works
- Base URL.
https://api.abusezero.com/v1/public/. Send noAuthorizationheader. - Same envelopes. Responses, lists and errors use the same shapes as the rest of the AbuseZero REST API. Every error uses the error envelope.
- Pages as Markdown. A docs page, help article, blog post or legal document comes with its
title,path,url,updated_atand itsmarkdown. - Cache validators. Every read returns an
ETag, and pages and lists also returnLast-Modified. Send them back asIf-None-MatchorIf-Modified-Since, and an unchanged answer returns304with no body. - Lists. Lists return
next_cursor. Pass it ascursorto read the next page. Cursors are opaque. - Limits. Reads have generous limits, and the contact and onboarding endpoints tight ones. Past a limit, a request returns
429withRATE_LIMITEDand aRetry-Afterheader. - Prices change. Read prices, credits and limits from the API each time you need them, instead of storing them.
The same reads work on the AbuseZero MCP server and in the AbuseZero CLI without signing in, and in the AbuseZero Node.js SDK through AbuseZeroPublic.
curl https://api.abusezero.com/v1/public/docs/quickstartimport { AbuseZeroPublic } from "abusezero/server";
const abusezero = new AbuseZeroPublic();
const page = await abusezero.docsPage("quickstart");Endpoints
| Method | Path | What it returns | MCP tool | Command |
|---|---|---|---|---|
GET | /v1/public/pricing | Plans, prices and what each plan includes | abusezero_get_pricing | pricing get |
GET | /v1/public/pricing/quote | The exact price of one choice | abusezero_get_price_quote | pricing quote |
GET | /v1/public/docs | The developer docs pages | abusezero_list_docs | docs list |
GET | /v1/public/docs/search | Docs pages that match a search | abusezero_search_docs | docs search |
GET | /v1/public/docs/{path} | A docs page as Markdown | abusezero_get_docs_page | docs get |
GET | /v1/public/help | The help center's articles | abusezero_list_help_articles | help-center list |
GET | /v1/public/help/search | Help articles that match a search | abusezero_search_help | help-center search |
GET | /v1/public/help/{path} | A help article as Markdown | abusezero_get_help_article | help-center get |
GET | /v1/public/openapi | The OpenAPI document | abusezero_get_openapi | openapi get |
GET | /v1/public/actions | Every action, with its MCP tool, command and scope | abusezero_list_actions | actions list |
GET | /v1/public/errors | Every error code | abusezero_list_error_codes | errors list |
GET | /v1/public/reasons | Every reason code | abusezero_list_reason_codes | reasons list |
GET | /v1/public/webhook-events | Webhook event types, with example payloads | abusezero_list_webhook_event_types | webhook-events list |
GET | /v1/public/conditions | The conditions custom policies can use | abusezero_list_conditions | conditions list |
GET | /v1/public/glossary | The glossary and the developer FAQ | abusezero_get_glossary | glossary get |
GET | /v1/public/legal | Legal documents | abusezero_list_legal_documents | legal list |
GET | /v1/public/legal/{document} | A legal document as Markdown | abusezero_get_legal_document | legal get |
GET | /v1/public/changelog | The product changelog | abusezero_get_changelog | changelog get |
GET | /v1/public/blog | Blog posts | abusezero_list_blog_posts | blog list |
GET | /v1/public/blog/{slug} | A blog post as Markdown | abusezero_get_blog_post | blog get |
GET | /v1/public/versions | The latest package version and browser script | abusezero_get_versions | versions get |
GET | /v1/public/status | Whether AbuseZero is operational | abusezero_get_service_status | service-status get |
GET | /v1/public/contacts | How to reach AbuseZero | abusezero_get_contacts | contacts get |
POST | /v1/public/contact/sales | A message to the sales team | abusezero_contact_sales | contact sales |
POST | /v1/public/contact/support | A request to support | abusezero_contact_support | contact support |
POST | /v1/public/onboarding-sessions | A new onboarding session | abusezero_start_onboarding | onboarding start |
POST | /v1/public/onboarding-sessions/{id}/status | Where an onboarding session stands | abusezero_get_onboarding_status | onboarding status |
Pricing
Read pricing
Also available as abusezero pricing get in the AbuseZero CLI, abusezero_get_pricing on the AbuseZero MCP server and pricing() in the AbuseZero Node.js SDK.
Returns every plan with its prices and what it includes, Builder's credit packs and what every plan shares. Builder is payg. Amounts are in US cents.
| Field | Type | Description |
|---|---|---|
credits_per_evaluation | integer | Credits one evaluation uses. |
plans[].plan | payg, pro or scale | The plan. |
plans[].name | string | The plan's name. |
plans[].prices.month, plans[].prices.year | object or null | amount_cents and the credits the period includes. null for Builder, which buys credit packs. |
plans[].includes | object | protected_benefits, team_seats, data_retention_days and traffic: rate_per_second, burst and concurrency for evaluations. |
credit_packs | object | credits_per_dollar, the credit packs' amounts in dollars, custom_amount_step, custom_amount_max and the first_purchase_multiplier. |
yearly | object | rollover_percent, the share of unused plan credits a yearly period converts at its end into credits that expire after converted_credit_expiry_months. |
every_plan | object | custom_policies, webhook_endpoints, webhook_resends_per_endpoint_per_hour and webhook_deliveries_per_day. |
enterprise.contact | string | Where to ask about an Enterprise plan. |
url | string | The pricing page. |
Get a price quote
Also available as abusezero pricing quote in the AbuseZero CLI, abusezero_get_price_quote on the AbuseZero MCP server and priceQuote(input) in the AbuseZero Node.js SDK.
The exact price of one choice for a new workspace's first purchase, worked out by AbuseZero. A credit pack includes the first-purchase bonus.
| Parameter | Type | Description |
|---|---|---|
plan | payg, pro or scale | Required. |
interval | month or year | Required for pro and scale. |
amount | integer | Required for payg: the credit pack's amount in US dollars. |
curl "https://api.abusezero.com/v1/public/pricing/quote?plan=pro&interval=year"The response is a price_quote with plan, interval, currency, amount_cents, credits, bonus_credits and total_credits. An amount the purchase can't take returns 422 with INVALID_FIELD and field: "amount".
Developer docs
List docs pages
Also available as abusezero docs list in the AbuseZero CLI, abusezero_list_docs on the AbuseZero MCP server and docs(query) in the AbuseZero Node.js SDK.
Returns a docs_page for every page, without its Markdown.
| Parameter | Type | Description |
|---|---|---|
limit | integer from 1 to 100 | Pages per response. Default 100. |
cursor | string | The next_cursor of the previous response. |
{
"object": "list",
"data": [
{
"object": "docs_page",
"path": "quickstart",
"title": "Quickstart",
"description": "Protect your first benefit with AbuseZero.",
"url": "https://docs.abusezero.com/docs/quickstart",
"updated_at": "2026-11-10T12:00:00.000Z"
}
],
"next_cursor": null
}The docs home page has the path index.
Search the docs
Also available as abusezero docs search in the AbuseZero CLI, abusezero_search_docs on the AbuseZero MCP server and searchDocs(query, limit) in the AbuseZero Node.js SDK.
Returns up to limit matches, best first. Every word in the query must appear in a page.
| Parameter | Type | Description |
|---|---|---|
query | string of 2 to 200 characters | Required. |
limit | integer from 1 to 25 | Default 10. |
Each match is a docs_page_match with path, title, url, a snippet and updated_at.
Read a docs page
Also available as abusezero docs get in the AbuseZero CLI, abusezero_get_docs_page on the AbuseZero MCP server and docsPage(path) in the AbuseZero Node.js SDK.
GET /v1/public/docs/{path}, where path is the page's path and may hold slashes, such as api-reference/errors. Returns the docs_page with its markdown. A path with no page returns 404 with PAGE_NOT_FOUND.
Help center
Also available as abusezero help-center list, abusezero help-center search and abusezero help-center get in the AbuseZero CLI, abusezero_list_help_articles, abusezero_search_help and abusezero_get_help_article on the AbuseZero MCP server and helpArticles(query), searchHelp(query, limit) and helpArticle(path) in the AbuseZero Node.js SDK.
These endpoints list articles, search their contents and return an article as Markdown. When the help center is unavailable, they return 404 with HELP_CENTER_UNAVAILABLE. Contact support remains available.
References
Read the OpenAPI document
Also available as abusezero openapi get in the AbuseZero CLI, abusezero_get_openapi on the AbuseZero MCP server and openapi() in the AbuseZero Node.js SDK.
Returns an openapi object with the OpenAPI version and the document, the machine-readable description of every action.
List actions
Also available as abusezero actions list in the AbuseZero CLI, abusezero_list_actions on the AbuseZero MCP server and actions() in the AbuseZero Node.js SDK.
Every action of the AbuseZero REST API, MCP server and CLI. Each action has its id, summary, method, path, scope, sign_in, mcp_tool, cli_command, mutation and destructive. Public actions have the scope public and sign_in: false.
List error codes
Also available as abusezero errors list in the AbuseZero CLI, abusezero_list_error_codes on the AbuseZero MCP server and errorCodes() in the AbuseZero Node.js SDK.
Every error code, each an error_code with its code, HTTP status, type, meaning, whether it's retryable, the next_step and its docs_url.
List reason codes
Also available as abusezero reasons list in the AbuseZero CLI, abusezero_list_reason_codes on the AbuseZero MCP server and reasonCodes() in the AbuseZero Node.js SDK.
Every reason code an evaluation returns, each a reason_code with its code, label, message, meaning and docs_url.
List webhook event types
Also available as abusezero webhook-events list in the AbuseZero CLI, abusezero_list_webhook_event_types on the AbuseZero MCP server and webhookEventTypes() in the AbuseZero Node.js SDK.
Every webhook event type, each a webhook_event_type with its type, a description and an example event in its exact shape. Example IDs belong to no record.
List conditions
Also available as abusezero conditions list in the AbuseZero CLI, abusezero_list_conditions on the AbuseZero MCP server and conditions() in the AbuseZero Node.js SDK.
The conditions the rules in custom policies can use, each a condition with its key, label, group and type, plus values, min, max or integer where they apply. Conditions on your own custom identifiers aren't listed here: list them for your workspace with GET /v1/policy-signals.
Read the glossary
Also available as abusezero glossary get in the AbuseZero CLI, abusezero_get_glossary on the AbuseZero MCP server and glossary() in the AbuseZero Node.js SDK.
Returns a glossary with every term and its definition, and the developer FAQ as faq: each entry's section, question and answer in Markdown.
Legal documents
Also available as abusezero legal list and abusezero legal get in the AbuseZero CLI, abusezero_list_legal_documents and abusezero_get_legal_document on the AbuseZero MCP server and legalDocuments() and legalDocument(document) in the AbuseZero Node.js SDK.
GET /v1/public/legal lists the Terms of Service, Privacy Policy, Cookie Policy, Acceptable Use Policy, Data Processing Agreement and Refund Policy. Each legal_document has its document, title, version, effective_date, url and updated_at. GET /v1/public/legal/{document} adds its markdown. document is terms, privacy, cookies, acceptable-use, dpa or refunds.
Read the changelog
Also available as abusezero changelog get in the AbuseZero CLI, abusezero_get_changelog on the AbuseZero MCP server and changelog() in the AbuseZero Node.js SDK.
Returns a changelog with its url and entries: each entry's id, title, date and markdown.
Blog
Also available as abusezero blog list and abusezero blog get in the AbuseZero CLI, abusezero_list_blog_posts and abusezero_get_blog_post on the AbuseZero MCP server and blogPosts(query) and blogPost(slug) in the AbuseZero Node.js SDK.
GET /v1/public/blog lists blog posts, newest first, with the same limit and cursor as the docs. Each blog_post has its slug, title, description, url, published_on, revised_on, authors, tags and updated_at. GET /v1/public/blog/{slug} adds its markdown.
Read the latest versions
Also available as abusezero versions get in the AbuseZero CLI, abusezero_get_versions on the AbuseZero MCP server and versions() in the AbuseZero Node.js SDK.
Returns a versions object: the API version, the npm package and its version, the browser_script with its version and url, and the AbuseZero skill with its version and url. The skill is null when the hosted skill is unavailable.
Read service status
Also available as abusezero service-status get in the AbuseZero CLI, abusezero_get_service_status on the AbuseZero MCP server and serviceStatus() in the AbuseZero Node.js SDK.
Returns a service_status with status, operational or degraded, and when it was checked_at.
Read contacts
Also available as abusezero contacts get in the AbuseZero CLI, abusezero_get_contacts on the AbuseZero MCP server and contacts() in the AbuseZero Node.js SDK.
Returns contacts: the support email, form and API endpoint, the sales form and its API endpoint, and the security email.
Contact sales
Also available as abusezero contact sales in the AbuseZero CLI, abusezero_contact_sales on the AbuseZero MCP server and contactSales(input, key) in the AbuseZero Node.js SDK.
Sends a message to AbuseZero's sales team about an Enterprise plan. Send an Idempotency-Key: a retry with the same key, email and message sends nothing new.
| Field | Type | Description |
|---|---|---|
name | string of 2 to 30 characters | Required. |
email | string | Required. Where the sales team replies. |
message | string of 10 to 380 characters | Required. |
Returns 201 with { "object": "sales_request", "received": true }.
Contact support
Also available as abusezero contact support in the AbuseZero CLI, abusezero_contact_support on the AbuseZero MCP server and contactSupport(input, key) in the AbuseZero Node.js SDK.
Sends a request to AbuseZero support without an account. The reply goes to the email you send. Send an Idempotency-Key: a retry with the same key and email returns the same reference.
| Field | Type | Description |
|---|---|---|
email | string | Required. Where support replies. |
subject | string of 1 to 120 characters | Required. |
message | string of 1 to 2,000 characters | Required. |
topic | account, billing, integration, dashboard or other | Required. |
name | string of 1 to 80 characters | Optional. |
workspace_id | string | Optional. The workspace the request is about. |
request_id | string of up to 100 characters | Optional. A request_id from an API response the request is about. |
Returns 201 with a support_request: its reference and the email support replies to. Reusing a key with different details returns 409 with IDEMPOTENCY_CONFLICT.
Agent onboarding
An agent can set up AbuseZero for someone who doesn't have an account yet. The agent agrees the plan and the workspace name with the person, starts an onboarding session and gives the person its link. The person reviews the details, changes any of them, signs in or creates an account, and pays. Then the agent collects a sign-in that reaches only the new workspace.
The quickest way is the AbuseZero CLI. It starts the session, opens the link, waits for the person and signs in once they've paid:
npx abusezero@latest setup --workspace-name "Acme" --plan pro- One link, for one person. The link works for 10 minutes until it's opened, then for an hour while the person signs in and pays. It works once, and the person can decline it.
- One secret, for the agent. Only the session's
secretcollects the sign-in. It's returned only when the session starts, so keep it to yourself and never show it to the person. - One workspace. The sign-in reaches only the workspace the session creates, with the scopes the person chose.
Start an onboarding session
Also available as abusezero onboarding start in the AbuseZero CLI, abusezero_start_onboarding on the AbuseZero MCP server and startOnboarding(input) in the AbuseZero Node.js SDK.
| Field | Type | Description |
|---|---|---|
workspace_name | string of 1 to 80 characters | Required. |
plan | payg, pro or scale | Required. payg is Builder. |
interval | month or year | Pro and Scale only. Leave it out for month. |
credit_pack_dollars | whole number | Builder only. The credit pack's amount in US dollars. Leave it out when the person hasn't named one, and AbuseZero's default pack applies. |
Returns 201 with an onboarding_session: its id, the approval_url to give the person, the secret, expires_at, the workspace_name, plan, interval and credit_pack_dollars it started with, and amount_cents, the first payment's price. The person can change any of these before paying.
Read an onboarding session's status
Also available as abusezero onboarding status in the AbuseZero CLI, abusezero_get_onboarding_status on the AbuseZero MCP server and onboardingStatus(id, input) in the AbuseZero Node.js SDK.
| Field | Type | Description |
|---|---|---|
secret | string | Required. The session's secret. |
wait | whole number of seconds, up to 25 | Optional. Waits up to this long for the person to finish before answering. |
Returns 200 with the session's status:
| Status | What it means |
|---|---|
waiting | The person hasn't finished yet. Read the status again. |
ready | The person paid. Collect the sign-in. |
collected | The sign-in was collected. |
declined | The person declined. Ask them before starting a new session. |
expired | The link expired. Start a new session if the person still wants to go ahead. |
With ready, workspace is the new workspace's ID, and sign_in says how to collect the sign-in: send grant_type, client_id, onboarding_session (the session's id) and session_secret (its secret) to token_endpoint as a form. It works once. The access token and refresh token work with the REST API and the MCP server, for the new workspace only.
An unknown id or a wrong secret returns 404 with ONBOARDING_SESSION_NOT_FOUND.
How is this guide?
Last updated on