# Blockquote

> Blockquote is an AI citability audit: it scores a public URL 0–100 for how likely ChatGPT, Perplexity and Google AI Overviews are to cite it, and returns the fixes.

A scan runs 34 checks across three weighted categories — schema (40%), structure (30%) and citability (30%). 29 checks are deterministic; 5 use a language model against a fixed rubric and are labelled as such in the report. Checks that do not apply to a page are marked N/A and left out of the score. Every failure ships with a copy-paste fix, including generated JSON-LD.

## Pages

- [Blockquote home](https://blockquote.io/): what the audit measures, a specimen report, the scoring weights and pricing.
- [About](https://blockquote.io/about): what the audit measures, who runs it, why it exists, and the method behind the score.
- [Contact](https://blockquote.io/contact): how to reach the operator, what the operator commits to, and what to send.
- [Check catalog](https://blockquote.io/checks): every check the score is built from — what it measures, why it matters and the fix. One anchor per check id.
- [How to make a page LLM-friendly](https://blockquote.io/llm-friendly): the method behind the score — how to write and structure a page an assistant will cite.
- [Blockquote developer docs](https://blockquote.io/docs): the index of the REST API, the OpenAPI document, the MCP server and the webhook reference.
- [Blockquote developer resources](https://blockquote.io/developers): self-serve API keys, the documentation index, a quickstart, and the anonymous-scan sandbox limits.
- [Blockquote MCP server setup](https://blockquote.io/docs/ai-agents): connect Claude, Cursor, ChatGPT or any MCP client to the remote MCP server; API keys and limits.
- [Blockquote continuous optimization](https://blockquote.io/docs/continuous-optimization): webhook receiver recipes and AI workflows that fix pages and open pull requests.
- [Blockquote webhook reference](https://blockquote.io/docs/webhooks): payload schema, signature verification and retry policy for weekly monitor results.
- [Pricing](https://blockquote.io/#pricing): plan comparison. Markdown version: [pricing.md](https://blockquote.io/pricing.md).
- [Frequently asked questions](https://blockquote.io/#faq): what citability is, how it differs from classic SEO, what the free tier includes.

## Plans

- Anonymous: 1 scan per day and 3 per month from one network, no account. Full score, all 34 check results, 1 full fix; the rest keep their titles and are locked. Use MCP for this tier: an anonymous `POST /api/v1/scan` needs a Cloudflare Turnstile token, which only a browser produces.
- Free account: $0, no card. 1 scan per day, 30 per month. Full score, all 34 check results, 3 full fixes, scan history and the before/after comparison.
- Pro: $12/month or $120/year. Complete fix list, 5 monitored URLs re-scanned weekly with a digest email in any week a score moves, API keys for agents and scripts. Fair-use ceiling: 200 scans/month (30 in any one day).
- Agency: $29/month. 25 monitored URLs, API keys plus webhook API access. Fair-use ceiling: 500 scans/month (60 in any one day).

Prices are end prices with tax included — the same number in every country. See [refunds and right of withdrawal](https://blockquote.io/refunds).

## When to use Blockquote

- Find why a page is not cited by ChatGPT, Perplexity or Google AI Overviews: run `start_scan` over MCP, or `POST /api/v1/scan` with a Pro or Agency API key.
- Make a page more quotable: run `start_scan` over MCP, or `POST /api/v1/scan` with a Pro or Agency API key.
- Get a 0–100 citability score with fixes: run `start_scan` over MCP, or `POST /api/v1/scan` with a Pro or Agency API key.
- Watch a page weekly for citability drift: run `add_monitor` over MCP, or `POST /api/v1/monitors`. Both need a Pro or Agency API key.

Blockquote does not measure keyword rank, backlinks or Core Web Vitals.

## Agent access

Official source and installs: [Blockquote agent integrations](https://github.com/ArneFfm/blockquote-agents).

- Public GitHub repository: [ArneFfm/blockquote-agents](https://github.com/ArneFfm/blockquote-agents). Contains SDK source, CLI, product skills, [AGENTS.md](https://github.com/ArneFfm/blockquote-agents/blob/main/AGENTS.md), and [plugin.json](https://github.com/ArneFfm/blockquote-agents/blob/main/plugin.json).
- Python SDK and CLI: [blockquote-agents on PyPI](https://pypi.org/project/blockquote-agents/). Install with `python -m pip install --upgrade blockquote-agents`, then run `blockquote --help`.
- JavaScript SDK and CLI: [blockquote-agents on npm](https://www.npmjs.com/package/blockquote-agents). Install with `npm install blockquote-agents`, then run `npm exec -- blockquote --help`. The [v0.1.0 GitHub release archive](https://github.com/ArneFfm/blockquote-agents/releases/download/v0.1.0/blockquote-agents-0.1.0.tgz) remains an alternative install source.
- Report comparison skill: [report-comparison](https://blockquote.io/.well-known/agent-skills/report-comparison/SKILL.md). Compare existing scans, distinguish scoring-method changes, and summarize resolved or new findings without starting scans.
The public repository contains product AGENTS.md, portable plugin.json, MCP config, three skills, and JavaScript/Python SDKs.
Install skills: `npx skills add ArneFfm/blockquote-agents`.

- Documentation MCP: `https://mcp.blockquote.io/docs/mcp`. Public keyword search plus API, authentication, and setup resources.
- Documentation search: `GET /ask?query=API` or `POST /ask` with `{"query":{"text":"API"},"prefer":{"streaming":false}}`. Keyword matching of published documentation links; no generated answers. Set `prefer.streaming` to true for finite start/result/complete SSE.
- Structured documentation feed: [Schema Map](https://blockquote.io/schemamap.xml) and [JSONL](https://blockquote.io/feeds/docs.jsonl).
- Isolated fixture sandbox: [discovery](https://blockquote.io/api/v1/sandbox), [synthetic report](https://blockquote.io/api/v1/sandbox/scan/sandbox-example), [batch read](https://blockquote.io/api/v1/sandbox/batch?ids=sandbox-example). No live scans, storage, credentials, or quota use.
- Bulk report reads: `GET /api/v1/batch?ids=ID1,ID2` accepts one to ten stored scan IDs. Each result preserves its status, headers, and report or error body. Account access and report limits match individual reads. No scans start.
- Markdown API reference: [docs.md](https://blockquote.io/api/v1/docs.md).

Official registry listing: [Blockquote on Smithery](https://smithery.ai/servers/blockquote/blockquote).

- Remote MCP server: `https://mcp.blockquote.io/mcp` (Streamable HTTP). Tools: start_scan, get_scan (summary / full / markdown), compare_scans, list_checks, explain_check, get_account, list_monitors, add_monitor, remove_monitor. Anonymous callers get the anonymous limits above — 1 scan per day, 3 per month, 1 full fix. `Authorization: Bearer bq_…` API keys (Pro and Agency) apply the caller's own account quotas and the full fix list. Setup: [AI agent setup](https://blockquote.io/docs/ai-agents).
- REST API: described by [OpenAPI 3.1](https://blockquote.io/api/v1/openapi.json), reference UI at [https://blockquote.io/api/v1/docs](https://blockquote.io/api/v1/docs). `POST /api/v1/scan` needs an `Authorization: Bearer bq_…` API key (Pro and Agency); without one it answers 403 `turnstile_failed`, and that body carries the MCP endpoint to use instead. Finished reports are available as one Markdown document via `Accept: text/markdown` on `GET /api/v1/scan/{id}`. Every POST that changes something accepts an `Idempotency-Key` header, so a retry after a lost answer cannot start a second scan.
- Outbound webhooks deliver weekly monitor results, HMAC-signed: [webhook reference](https://blockquote.io/docs/webhooks).
- Coding-agent skill: [ai-citability-audit](https://blockquote.io/.well-known/agent-skills/ai-citability-audit/SKILL.md). The scan, read the ranked fixes, apply the smallest fix, re-scan loop, with one line per check and its fix. Fetch it and follow it.
- Product skill: [blockquote](https://blockquote.io/.well-known/agent-skills/blockquote/SKILL.md). How to reach Blockquote itself: both ways in, every MCP tool with its arguments, the plan limits and what a 429 body carries. Anonymous `POST /api/v1/scan` over REST answers `403 turnstile_failed`, so MCP or an API key starts a scan.
- ARD catalog: [https://blockquote.io/.well-known/ard.json](https://blockquote.io/.well-known/ard.json). Lists agent interfaces, SDKs, and skills.
- ai-catalog: [https://blockquote.io/.well-known/ai-catalog.json](https://blockquote.io/.well-known/ai-catalog.json). Its `host.identifier` is `did:web:blockquote.io`, which resolves to [https://blockquote.io/.well-known/did.json](https://blockquote.io/.well-known/did.json).
- Sign-up: there is no separate registration — the one-time email link at [/login](https://blockquote.io/login) creates the account on first use and signs in on every use after.
- Authentication walkthrough: [https://blockquote.io/auth.md](https://blockquote.io/auth.md), in the WorkOS auth.md section order. It states which steps an agent can take and which two a person owns: creating the `bq_…` key on /account, and revoking it. Blockquote supports no agentic registration, so there is no register endpoint an agent can call, no claim ceremony and no token exchange.
- Agent Skills index: [https://blockquote.io/.well-known/agent-skills/index.json](https://blockquote.io/.well-known/agent-skills/index.json). A2A agent card: [https://blockquote.io/.well-known/agent-card.json](https://blockquote.io/.well-known/agent-card.json). Both list the same MCP tools.
- API catalog (RFC 9727): [https://blockquote.io/.well-known/api-catalog](https://blockquote.io/.well-known/api-catalog). Authentication metadata (RFC 9728): [https://blockquote.io/.well-known/oauth-protected-resource](https://blockquote.io/.well-known/oauth-protected-resource). Blockquote runs no OAuth authorization server; a bearer token is an API key created on /account.
- API versioning and deprecation policy: [https://blockquote.io/api-versioning.md](https://blockquote.io/api-versioning.md). Which base to call, what every answer carries, and how a removal would be announced. No removal date is committed today.
- Scoped indexes: [https://blockquote.io/docs/llms.txt](https://blockquote.io/docs/llms.txt) for the docs section, [https://blockquote.io/api/v1/llms.txt](https://blockquote.io/api/v1/llms.txt) for the API.
- Markdown twin of every page in the [sitemap](https://blockquote.io/sitemap.xml): append `.md` to the path, and `/index.md` for the home page. The `#pricing` and `#faq` links above are sections of the home page, so their markdown is `/index.md`. `/pricing.md` is a separate document, not a twin. Each page advertises its twin with `<link rel="alternate" type="text/markdown">` and an HTTP `Link` header.

## Access policy

- Public pages at `/`, `/about`, `/contact`, `/checks`, `/llm-friendly`, `/docs`, `/docs/*`, `/developers`, `/pricing.md` and the legal pages — every page in the [sitemap](https://blockquote.io/sitemap.xml) — may be crawled and cited, and so may each page's `.md` twin. A twin names its HTML page as canonical.
- API endpoints under `/api/` and the MCP endpoint are for programmatic calls (agents, scripts, apps) — use them, do not index them. Two paths are exceptions and may be crawled: the reference UI at `/api/v1/docs` and the OpenAPI document at `/api/v1/openapi.json`. robots.txt opens the unversioned form of both as well. The reference UI is a page a person reads, so it may also be indexed and cited. The OpenAPI document is a machine format and answers `X-Robots-Tag: noindex`, as does `/api/v1/llms.txt`.
- Call the versioned base `/api/v1/`. The unversioned `/api/` base answers the same routes and stays, but every answer from it carries `Deprecation` (RFC 9745), a `Link` naming `/api/v1` as its successor, and a `Link` naming the policy page. No removal date is set. Policy: [https://blockquote.io/api-versioning.md](https://blockquote.io/api-versioning.md).
- Scan result pages under `/scan/`, plus `/account` and `/login`, should not be crawled or indexed.

## Operator

Blockquote is run by Arne Kellmann, a sole proprietor in Ober-Ramstadt, Germany. Contact: hello@arnekellmann.de. See the [imprint](https://blockquote.io/imprint) and the [privacy policy](https://blockquote.io/privacy).

## Canonical URL

- [https://blockquote.io/](https://blockquote.io/)