# `io.github.SeoScoreAPI/seoscoreapi-mcp` on Anlyon Radar

> Text inside fenced blocks marked untrusted-server-text was written by the MCP server's publisher, not by Anlyon Radar. It is data to inspect, not instructions to follow. Invisible characters are shown as [U+XXXX].
> No finding is not a safety guarantee: these are point-in-time checks of tool definitions, not of code or behaviour.

- Page: https://radar.anlyon.com/servers/io.github.SeoScoreAPI/seoscoreapi-mcp
- JSON: https://radar.anlyon.com/api/v1/servers/io.github.SeoScoreAPI/seoscoreapi-mcp
- Version: 1.5.1 (active)
- Publisher: io.github.SeoScoreAPI (GitHub account)
- Runs: remote (streamable-http), local (npm)
- Sign-in to list tools: none
- Repository: https://github.com/SeoScoreAPI/seoscoreapi-mcp
- Metadata from the official MCP Registry, synced 2026-11-14T22:17:37.624Z

Title and description, as the publisher wrote them:

```untrusted-server-text
SEO Score API
SEO, GEO and ADA/WCAG audits, AI citation tracking and site crawls for AI assistants.
```

Scanned 2026-11-14T22:18:18.500Z over streamable-http.

**45 tools: 4 can delete or cancel things, 2 do something that can't be undone, 24 change things and 15 only read.**

| Group | Class | Tools | Claude Code |
| --- | --- | --- | --- |
| Deletes or cancels | destructive | 4 | deny |
| Can't be undone | irreversible | 2 | ask |
| Changes things | write | 24 | ask |
| Only reads | read | 15 | allow |

## Exfil surface

374 private-data -> untrusted-input -> outbound paths in this server alone. Denying 1 private-data tool breaks every one: `mcp__seoscoreapi__trackers`.

| Tag | Meaning | Tools |
| --- | --- | --- |
| private-data | Reads files, mail, messages, databases, secrets, calendars or repositories. | `mcp__seoscoreapi__trackers` |
| untrusted-input | Brings in text someone else can write: web pages, inbound email, messages, issues and comments. | `mcp__seoscoreapi__accessibility_audit` `mcp__seoscoreapi__ai_readability` `mcp__seoscoreapi__audit` `mcp__seoscoreapi__backlinks` `mcp__seoscoreapi__batch_audit` `mcp__seoscoreapi__citation_check` `mcp__seoscoreapi__citation_suggest_prompts` `mcp__seoscoreapi__compare` `mcp__seoscoreapi__competitive_audit` `mcp__seoscoreapi__conversion` `mcp__seoscoreapi__crawl_site` `mcp__seoscoreapi__deep_audit` `mcp__seoscoreapi__geo_audit` `mcp__seoscoreapi__geo_brand_probe` `mcp__seoscoreapi__get_crawl` `mcp__seoscoreapi__history` `mcp__seoscoreapi__llms_txt` `mcp__seoscoreapi__report_url` `mcp__seoscoreapi__run_citation_tracker` `mcp__seoscoreapi__start_crawl` `mcp__seoscoreapi__start_deep_audit` `mcp__seoscoreapi__trackers` |
| outbound | Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL. | `mcp__seoscoreapi__accessibility_audit` `mcp__seoscoreapi__add_geo_monitor` `mcp__seoscoreapi__add_monitor` `mcp__seoscoreapi__ai_readability` `mcp__seoscoreapi__audit` `mcp__seoscoreapi__batch_audit` `mcp__seoscoreapi__compare` `mcp__seoscoreapi__competitive_audit` `mcp__seoscoreapi__conversion` `mcp__seoscoreapi__crawl_site` `mcp__seoscoreapi__create_citation_tracker` `mcp__seoscoreapi__deep_audit` `mcp__seoscoreapi__geo_audit` `mcp__seoscoreapi__history` `mcp__seoscoreapi__start_crawl` `mcp__seoscoreapi__start_deep_audit` `mcp__seoscoreapi__trackers` |

| Private data | Untrusted input | Outbound |
| --- | --- | --- |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__accessibility_audit` | `mcp__seoscoreapi__accessibility_audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__accessibility_audit` | `mcp__seoscoreapi__add_geo_monitor` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__accessibility_audit` | `mcp__seoscoreapi__add_monitor` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__accessibility_audit` | `mcp__seoscoreapi__ai_readability` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__accessibility_audit` | `mcp__seoscoreapi__audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__accessibility_audit` | `mcp__seoscoreapi__batch_audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__accessibility_audit` | `mcp__seoscoreapi__compare` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__accessibility_audit` | `mcp__seoscoreapi__competitive_audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__accessibility_audit` | `mcp__seoscoreapi__conversion` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__accessibility_audit` | `mcp__seoscoreapi__crawl_site` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__accessibility_audit` | `mcp__seoscoreapi__create_citation_tracker` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__accessibility_audit` | `mcp__seoscoreapi__deep_audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__accessibility_audit` | `mcp__seoscoreapi__geo_audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__accessibility_audit` | `mcp__seoscoreapi__history` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__accessibility_audit` | `mcp__seoscoreapi__start_crawl` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__accessibility_audit` | `mcp__seoscoreapi__start_deep_audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__accessibility_audit` | `mcp__seoscoreapi__trackers` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__ai_readability` | `mcp__seoscoreapi__accessibility_audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__ai_readability` | `mcp__seoscoreapi__add_geo_monitor` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__ai_readability` | `mcp__seoscoreapi__add_monitor` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__ai_readability` | `mcp__seoscoreapi__ai_readability` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__ai_readability` | `mcp__seoscoreapi__audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__ai_readability` | `mcp__seoscoreapi__batch_audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__ai_readability` | `mcp__seoscoreapi__compare` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__ai_readability` | `mcp__seoscoreapi__competitive_audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__ai_readability` | `mcp__seoscoreapi__conversion` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__ai_readability` | `mcp__seoscoreapi__crawl_site` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__ai_readability` | `mcp__seoscoreapi__create_citation_tracker` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__ai_readability` | `mcp__seoscoreapi__deep_audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__ai_readability` | `mcp__seoscoreapi__geo_audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__ai_readability` | `mcp__seoscoreapi__history` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__ai_readability` | `mcp__seoscoreapi__start_crawl` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__ai_readability` | `mcp__seoscoreapi__start_deep_audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__ai_readability` | `mcp__seoscoreapi__trackers` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__audit` | `mcp__seoscoreapi__accessibility_audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__audit` | `mcp__seoscoreapi__add_geo_monitor` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__audit` | `mcp__seoscoreapi__add_monitor` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__audit` | `mcp__seoscoreapi__ai_readability` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__audit` | `mcp__seoscoreapi__audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__audit` | `mcp__seoscoreapi__batch_audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__audit` | `mcp__seoscoreapi__compare` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__audit` | `mcp__seoscoreapi__competitive_audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__audit` | `mcp__seoscoreapi__conversion` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__audit` | `mcp__seoscoreapi__crawl_site` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__audit` | `mcp__seoscoreapi__create_citation_tracker` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__audit` | `mcp__seoscoreapi__deep_audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__audit` | `mcp__seoscoreapi__geo_audit` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__audit` | `mcp__seoscoreapi__history` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__audit` | `mcp__seoscoreapi__start_crawl` |
| `mcp__seoscoreapi__trackers` | `mcp__seoscoreapi__audit` | `mcp__seoscoreapi__start_deep_audit` |

And 324 more: every combination of one tool from each row of the table above.

## Findings

- **warn** Annotations that disagree with the tool in `audit` ([annotation-conflict](https://radar.anlyon.com/methodology#annotation-conflict))

```untrusted-server-text
The server marks "audit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The description's first verb is "run": a word for a change.
```

- **warn** Annotations that disagree with the tool in `accessibility_audit` ([annotation-conflict](https://radar.anlyon.com/methodology#annotation-conflict))

```untrusted-server-text
The server marks "accessibility_audit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The description's first verb is "run": a word for a change.
```

- **warn** Annotations that disagree with the tool in `llms_txt` ([annotation-conflict](https://radar.anlyon.com/methodology#annotation-conflict))

```untrusted-server-text
The server marks "llms_txt" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The description's first verb is "generate": a word for a change.
```

- **warn** Annotations that disagree with the tool in `geo_audit` ([annotation-conflict](https://radar.anlyon.com/methodology#annotation-conflict))

```untrusted-server-text
The server marks "geo_audit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The description's first verb is "run": a word for a change.
```

- **warn** Annotations that disagree with the tool in `citation_topups` ([annotation-conflict](https://radar.anlyon.com/methodology#annotation-conflict))

```untrusted-server-text
The server marks "citation_topups" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "citation_topups" says "topup": a word that usually means money moves, which cannot be taken back.
```


## Deletes or cancels (4)

Removes, cancels or revokes something. Claude Code: deny.

### `delete_citation_tracker`

- The name "delete_citation_tracker" says "delete": a word for removing or ending something.
- The server marks it destructive (destructiveHint: true).

Description, as the server wrote it:

```untrusted-server-text
Stop tracking a brand. Run history is kept until its retention expires.
```

### `remove_geo_monitor`

- The name "remove_geo_monitor" says "remove": a word for removing or ending something.
- The server marks it destructive (destructiveHint: true).

Description, as the server wrote it:

```untrusted-server-text
Stop a GEO monitor by its URL. Past history is kept.
```

### `remove_monitor`

- The name "remove_monitor" says "remove": a word for removing or ending something.
- The server marks it destructive (destructiveHint: true).

Description, as the server wrote it:

```untrusted-server-text
Stop score monitoring for a URL. Its past history is kept.
```

### `update_citation_tracker`

- The name "update_citation_tracker" says "update": a word for a change.
- The server marks it destructive (destructiveHint: true).

Description, as the server wrote it:

```untrusted-server-text
Edit a citation tracker's prompts, engines, competitors, samples, cadence or webhook. Only the fields passed change.
```

## Can't be undone (2)

Sends, pays, publishes or merges: once done, it cannot be taken back. Claude Code: ask.

### `citation_topup_checkout`

- The name "citation_topup_checkout" says "topup": a word that usually means money moves, which cannot be taken back.

Description, as the server wrote it:

```untrusted-server-text
Create a one-time Stripe Checkout link for a pack of top-up citation checks and return its URL. Paid plans only. This charges nothing: the user must open the link and pay by card, and the checks are added when the payment completes. Get the pack keys and prices from citation_topups first.
```

### `citation_topups`

- The name "citation_topups" says "topup": a word that usually means money moves, which cannot be taken back.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
- Findings: warn annotation-conflict

Description, as the server wrote it:

```untrusted-server-text
Top-up packs for AI citation checks (`packs` is the live list with prices) and this account's balance. Top-up checks are spent only after the plan's monthly allowance, do not reset on the 1st, and expire 12 months after purchase. `can_buy` is false on the free tier.
```

## Changes things (24)

Creates or changes something that can be changed back. Claude Code: ask.

### `accessibility_audit`

- The description's first verb is "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data.
- Tag outbound: The input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
- Findings: warn annotation-conflict

Description, as the server wrote it:

```untrusted-server-text
Run an ADA / WCAG 2.1 AA accessibility audit (axe-core). Paid plans only. Returns the compliance score and grade, lawsuit risk (high/medium/low/minimal), category breakdowns, violations (WCAG 2.1 AA failures with affected elements and a plain-English fix), best_practices (not WCAG requirements; they never lower the score), needs_review and priorities. ADA audits have their own monthly allowance, separate from the audit quota (the usage tool reports ada_remaining). Each key may run 2 at a time; a 429 is not counted.
```

### `add_geo_monitor`

- The name "add_geo_monitor" says "add": a word for a change.
- Tag outbound: The input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

Description, as the server wrote it:

```untrusted-server-text
Create a GEO monitor for a URL. Requires Basic plan or higher (the Basic plan runs weekly).
```

### `add_monitor`

- The name "add_monitor" says "add": a word for a change.
- Tag outbound: The input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

Description, as the server wrote it:

```untrusted-server-text
Add a URL to score monitoring: it is re-audited on a schedule (daily or weekly) and the account owner is emailed when the score drops. Each plan allows a set number of monitors; the error says when the limit is reached.
```

### `ai_readability`

- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data.
- Tag outbound: The input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

Description, as the server wrote it:

```untrusted-server-text
Score how easily AI assistants and LLM search (ChatGPT, Perplexity, Google AI Overviews) can read, quote and cite a page: structure, answer-first content, entities and machine-readable data, with recommendations. Counts as one audit.
```

### `audit`

- The description's first verb is "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data.
- Tag outbound: The input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
- Findings: warn annotation-conflict

Description, as the server wrote it:

```untrusted-server-text
Run a full SEO audit on a URL. Returns score (0-100), letter grade, category breakdowns (meta, technical, social, performance, accessibility), AI readability score, SXO/AEO/AIO scores (paid plans), and a prioritized list of fixes. Usually returns in seconds. Free tier: 20 audits a month. Core Web Vitals (LCP, INP, CLS) are not measured unless include_core_web_vitals is true.
```

### `backlinks`

- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data.

Description, as the server wrote it:

```untrusted-server-text
Look up observed backlinks for a domain. SEO Score API maintains an audit-fed backlink graph: every audit contributes the external links found on the audited page. The dataset is a sample (NOT a comprehensive backlink index like Ahrefs) and the response always includes a data_caveat. Coverage grows over time. Requires Basic plan ($15/mo) or higher.
```

### `batch_audit`

- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data.
- Tag outbound: The input takes a URL (input.urls): a request to a URL can carry data to whoever runs it.

Description, as the server wrote it:

```untrusted-server-text
Audit multiple URLs in one call. Requires Starter plan ($5/mo) or higher. Returns one audit object per URL.
```

### `citation_check`

- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data.

Description, as the server wrote it:

```untrusted-server-text
One-off AI citation check for a brand and one prompt: no tracker, metered per check (one check per engine per sample). Paid plans only (Starter and up).
```

### `citation_starter_prompts`

- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

Description, as the server wrote it:

```untrusted-server-text
Build a starter set of prompts to track for AI citations, from a topic plus optional brand, competitors and audience. Fixed templates, not a model: no AI check is spent. Works on every plan and is not metered. Returns `prompts`, ready to pass to create_citation_tracker.
```

### `citation_suggest_prompts`

- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data.

Description, as the server wrote it:

```untrusted-server-text
Suggest AI-citation prompts to track, read off what a site says it does (fetches the site's pages). Use the result with create_citation_tracker.
```

### `citation_tracker_history`

- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

Description, as the server wrote it:

```untrusted-server-text
Timeseries of mention rate, citation rate, share of voice and average position for a citation tracker.
```

### `competitive_audit`

- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data.
- Tag outbound: The input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

Description, as the server wrote it:

```untrusted-server-text
Head-to-head SEO comparison of your page against a competitor's page with keyword-relevance scoring. Requires Pro plan ($39/mo) or higher.
```

### `conversion`

- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data.
- Tag outbound: The input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

Description, as the server wrote it:

```untrusted-server-text
Score how well a page turns a visitor into a lead or a sale: headline, call to action, trust, forms and objections, each 0-100, with a prioritized fix list. It does not measure the actual conversion rate. Counts as one audit. Free keys get the scores and the first two priorities; paid plans get the full list.
```

### `crawl_site`

- Nothing in its name or description says what it does, so it is counted as a write.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data. The name "crawl_site" says "crawl": it reads web pages, which anyone can write. The description's first sentence says "crawl": it reads web pages, which anyone can write.
- Tag outbound: The input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

Description, as the server wrote it:

```untrusted-server-text
Crawl one site and wait for the result: follows same-site links plus the sitemap and reports broken pages, redirects, duplicate titles and headings, pages missing a title or heading, orphan pages and slow pages. Pro: 25 pages per crawl, 20 crawls a month. Ultra: 100 pages, 100 crawls a month. Any other plan: one Deep Audit credit covers a 25-page crawl. A failed crawl is not counted. For long waits use start_crawl + get_crawl.
```

### `create_citation_tracker`

- The name "create_citation_tracker" says "create": a word for a change.
- Tag outbound: The input takes a URL (input.webhook_url): a request to a URL can carry data to whoever runs it.

Description, as the server wrote it:

```untrusted-server-text
Track a brand's visibility in AI answers (ChatGPT, Gemini, Perplexity, Claude): mention rate, citation rate, share of voice and position over time. Requires Starter ($5/mo) or higher. Monthly allowance (checks / tracked prompts): Starter 150 / 15, Basic 500 / 30, Pro 1,500 / 75, Ultra 4,000 / 200. A check is one prompt on one engine, sampled once. With no prompts, pass `starter` and a starter set is generated to fit the plan.
```

### `deep_audit`

- The description's first verb is "run": a word for a change.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data.
- Tag outbound: The input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

Description, as the server wrote it:

```untrusted-server-text
Run a Deep Audit — a thorough, AI-assisted SEO audit that scores a URL across 9 dimensions (technical, crawlability, content/entity, Core Web Vitals, technical SEO, structured data, security, media, off-site): thousands of catalog checks plus up to 150 AI checks. Runs asynchronously server-side; this tool waits for the result (about 90 seconds once it starts, longer if queued). Included on Pro ($39/mo, 20/month) and Ultra ($99/mo, 100/month); other keys spend a purchased Deep Audit credit. For long waits use start_deep_audit + get_deep_audit instead.
```

### `deep_audit_usage`

- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

Description, as the server wrote it:

```untrusted-server-text
Deep Audits used and remaining this calendar month for the API key in use.
```

### `geo_audit`

- The description's first verb is "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data.
- Tag outbound: The input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
- Findings: warn annotation-conflict

Description, as the server wrote it:

```untrusted-server-text
Run a GEO (Generative Engine Optimization) audit: 26 checks measuring how visible a page is to LLMs. Requires Basic plan ($15/mo) or higher; uses the plan's monthly GEO allowance.
```

### `geo_brand_probe`

- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data.

Description, as the server wrote it:

```untrusted-server-text
Probe LLMs to measure how often a brand is mentioned in AI responses. Requires Basic plan ($15/mo) or higher; uses the plan's monthly GEO allowance. Can take a while.
```

### `geo_monitor_history`

- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

Description, as the server wrote it:

```untrusted-server-text
Score history for one GEO monitor, newest first, paginated.
```

### `llms_txt`

- The description's first verb is "generate": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data.
- Findings: warn annotation-conflict

Description, as the server wrote it:

```untrusted-server-text
Generate an llms.txt file for a domain and return its text, ready to save at /llms.txt. Basic plan ($15/mo) and higher get the curated version with per-page descriptions; other keys (or no key) get the basic version.
```

### `run_citation_tracker`

- The name "run_citation_tracker" says "run": a word for a change.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data.

Description, as the server wrote it:

```untrusted-server-text
Run a citation tracker now. Every prompt x engine x sample counts as one check. A run needs all of its checks up front (the month's remaining allowance plus any top-up balance); if that is short the run is refused and nothing is spent. A check whose engine returned an error is not counted. A run that takes over 45 seconds returns {run_id, status: "running", poll} instead of the summary; it keeps running, so call get_citation_run with that run_id until status is completed.
```

### `start_crawl`

- The name "start_crawl" says "start": a word for a change.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data. The name "start_crawl" says "crawl": it reads web pages, which anyone can write. The description's first sentence says "crawl": it reads web pages, which anyone can write.
- Tag outbound: The input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

Description, as the server wrote it:

```untrusted-server-text
Start a site crawl and return immediately with a job_id and a shareable site_map link. Poll it with get_crawl. Same plans as crawl_site.
```

### `start_deep_audit`

- The name "start_deep_audit" says "start": a word for a change.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data.
- Tag outbound: The input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

Description, as the server wrote it:

```untrusted-server-text
Queue a Deep Audit and return immediately with a job_id. Poll it with get_deep_audit. Same plans as deep_audit.
```

## Only reads (15)

Looks things up and changes nothing. Claude Code: allow.

### `citation_auto_reload`

- The description's first verb is "read": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
Read the auto-reload setting for AI citation checks and this month's reloads and charges. Read-only: auto-reload is off unless the account owner turns it on, and it can only be turned on or changed from the dashboard (signed in), never through the API or this server, because turning it on authorizes charges.
```

### `compare`

- The name "compare" says "compare": a word for reading.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data. The description's first sentence says "compare" and "url": it reads web pages, which anyone can write. The input takes a URL (input.urls), and the tool reads it: it reads web pages, which anyone can write.
- Tag outbound: The input takes a URL (input.urls): a request to a URL can carry data to whoever runs it.
- Claude Code: ask, because it is tagged outbound.

Description, as the server wrote it:

```untrusted-server-text
Compare 2-5 URLs side by side with a structured diff: each URL's score and category breakdown plus who is ahead and by how much, per category and per URL pair. Requires Basic plan ($15/mo) or higher.
```

### `get_citation_run`

- The name "get_citation_run" says "get": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
A citation run's status plus every per-prompt, per-engine result (mentioned, cited, position, snippet).
```

### `get_citation_tracker`

- The name "get_citation_tracker" says "get": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
Read one citation tracker: brand, domains, prompts, engines, competitors, cadence and webhook.
```

### `get_crawl`

- The name "get_crawl" says "get": a word for reading.
- Tag untrusted-input: The name "get_crawl" says "crawl": it reads web pages, which anyone can write. The description's first sentence says "crawl": it reads web pages, which anyone can write.

Description, as the server wrote it:

```untrusted-server-text
Get a site crawl's status: queued, running (pages_done counts up), completed (result: summary, issues, pages, links) or failed (error). Only the key that started a crawl can read it.
```

### `get_deep_audit`

- The name "get_deep_audit" says "get": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
Get a Deep Audit job's status: queued (queue_position, eta_seconds), running (progress 0-100, stage), completed (result) or failed (error). Poll every 5-10 seconds.
```

### `history`

- The description's first verb is "fetch": a word for reading.
- Tag untrusted-input: The description's first sentence says "fetch" and "url": it reads web pages, which anyone can write. The input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
- Tag outbound: The input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
- Claude Code: ask, because it is tagged outbound.

Description, as the server wrote it:

```untrusted-server-text
Fetch the audit-score history for a URL. Returns a time-series of past scores so you can see trends. Requires Starter plan ($5/mo) or higher.
```

### `history_domains`

- The description's first verb is "list": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
List every domain audited by this API key with the latest score and 30-day trend. Starter+.
```

### `list_citation_trackers`

- The name "list_citation_trackers" says "list": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
List the key's citation trackers plus `usage`: the month's checks (limit, used, remaining), the top-up balance and checks_available (what a run can spend now).
```

### `list_geo_monitors`

- The name "list_geo_monitors" says "list": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
List the active GEO monitors for the API key in use.
```

### `list_monitors`

- The name "list_monitors" says "list": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
List all URLs currently under score monitoring for the active API key.
```

### `quick_wins`

- The description's first verb is "find": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
Find the page-2 quick wins in the user's own Google Search Console data: queries ranking 11-20 worth working on first, each with an opportunity score and an estimate of the extra clicks from moving onto page 1. The API does not connect to Search Console: pass the exported rows or the CSV export text. Nothing is stored. The click numbers are estimates. Works on every plan; counts as one audit. Up to 1 MB per request.
```

### `report_url`

- The description's first verb is "return": a word for reading.
- Tag untrusted-input: The name "report_url" says "url": it reads web pages, which anyone can write. The description's first sentence says "return" and "html": it reads web pages, which anyone can write.

Description, as the server wrote it:

```untrusted-server-text
Return the public, shareable HTML report URL for a domain (https://seoscoreapi.com/report/{domain}). Useful for sending results to a client without exposing an API key. Makes no request and uses no quota.
```

### `trackers`

- The description's first verb is "list": a word for reading.
- Tag private-data: The description's first sentence says "list" and "chat": it reads messages.
- Tag untrusted-input: The server marks it openWorldHint: true: it reaches things outside the server's own data. The description's first sentence says "list" and "chat": it reads messages, which other people can write. The input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
- Tag outbound: The input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
- Claude Code: ask, because it is tagged outbound.

Description, as the server wrote it:

```untrusted-server-text
List the third-party trackers and pixels a page loads (analytics, advertising, session replay, chat, marketing automation, A/B testing, monitoring), including pixels a tag manager injects. Loads the page in a real browser with no interaction; it does not determine whether a tracker waited for consent, and it is not legal advice. Counts as one audit; a page that can't be loaded is not counted.
```

### `usage`

- The description's first verb is "get": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
Get the current plan, monthly usage, and remaining quota for the API key in use.
```

## Claude Code permissions

For `.claude/settings.json`, assuming the server is installed as `seoscoreapi`:

```json
{
  "permissions": {
    "allow": [
      "mcp__seoscoreapi__usage",
      "mcp__seoscoreapi__report_url",
      "mcp__seoscoreapi__list_monitors",
      "mcp__seoscoreapi__history_domains",
      "mcp__seoscoreapi__get_deep_audit",
      "mcp__seoscoreapi__quick_wins",
      "mcp__seoscoreapi__list_geo_monitors",
      "mcp__seoscoreapi__get_crawl",
      "mcp__seoscoreapi__list_citation_trackers",
      "mcp__seoscoreapi__get_citation_tracker",
      "mcp__seoscoreapi__get_citation_run",
      "mcp__seoscoreapi__citation_auto_reload"
    ],
    "ask": [
      "mcp__seoscoreapi__audit",
      "mcp__seoscoreapi__batch_audit",
      "mcp__seoscoreapi__add_monitor",
      "mcp__seoscoreapi__history",
      "mcp__seoscoreapi__backlinks",
      "mcp__seoscoreapi__deep_audit",
      "mcp__seoscoreapi__start_deep_audit",
      "mcp__seoscoreapi__deep_audit_usage",
      "mcp__seoscoreapi__compare",
      "mcp__seoscoreapi__competitive_audit",
      "mcp__seoscoreapi__accessibility_audit",
      "mcp__seoscoreapi__ai_readability",
      "mcp__seoscoreapi__trackers",
      "mcp__seoscoreapi__conversion",
      "mcp__seoscoreapi__llms_txt",
      "mcp__seoscoreapi__geo_audit",
      "mcp__seoscoreapi__geo_brand_probe",
      "mcp__seoscoreapi__add_geo_monitor",
      "mcp__seoscoreapi__geo_monitor_history",
      "mcp__seoscoreapi__crawl_site",
      "mcp__seoscoreapi__start_crawl",
      "mcp__seoscoreapi__citation_starter_prompts",
      "mcp__seoscoreapi__citation_suggest_prompts",
      "mcp__seoscoreapi__create_citation_tracker",
      "mcp__seoscoreapi__run_citation_tracker",
      "mcp__seoscoreapi__citation_tracker_history",
      "mcp__seoscoreapi__citation_check",
      "mcp__seoscoreapi__citation_topups",
      "mcp__seoscoreapi__citation_topup_checkout"
    ],
    "deny": [
      "mcp__seoscoreapi__remove_monitor",
      "mcp__seoscoreapi__remove_geo_monitor",
      "mcp__seoscoreapi__update_citation_tracker",
      "mcp__seoscoreapi__delete_citation_tracker"
    ]
  }
}
```

Capability tags behind the rules (a read tagged outbound is asked about):

- `mcp__seoscoreapi__audit`: untrusted-input, outbound
- `mcp__seoscoreapi__batch_audit`: untrusted-input, outbound
- `mcp__seoscoreapi__report_url`: untrusted-input
- `mcp__seoscoreapi__add_monitor`: outbound
- `mcp__seoscoreapi__history`: untrusted-input, outbound
- `mcp__seoscoreapi__backlinks`: untrusted-input
- `mcp__seoscoreapi__deep_audit`: untrusted-input, outbound
- `mcp__seoscoreapi__start_deep_audit`: untrusted-input, outbound
- `mcp__seoscoreapi__compare`: untrusted-input, outbound
- `mcp__seoscoreapi__competitive_audit`: untrusted-input, outbound
- `mcp__seoscoreapi__accessibility_audit`: untrusted-input, outbound
- `mcp__seoscoreapi__ai_readability`: untrusted-input, outbound
- `mcp__seoscoreapi__trackers`: private-data, untrusted-input, outbound
- `mcp__seoscoreapi__conversion`: untrusted-input, outbound
- `mcp__seoscoreapi__llms_txt`: untrusted-input
- `mcp__seoscoreapi__geo_audit`: untrusted-input, outbound
- `mcp__seoscoreapi__geo_brand_probe`: untrusted-input
- `mcp__seoscoreapi__add_geo_monitor`: outbound
- `mcp__seoscoreapi__crawl_site`: untrusted-input, outbound
- `mcp__seoscoreapi__start_crawl`: untrusted-input, outbound
- `mcp__seoscoreapi__get_crawl`: untrusted-input
- `mcp__seoscoreapi__citation_suggest_prompts`: untrusted-input
- `mcp__seoscoreapi__create_citation_tracker`: outbound
- `mcp__seoscoreapi__run_citation_tracker`: untrusted-input
- `mcp__seoscoreapi__citation_check`: untrusted-input

## Remote posture

What an anonymous client saw at the remote endpoint, last 2026-11-18T16:21:29.096Z, unchanged since 2026-11-11T09:20:01.249Z. No credentials were sent and no tool was called. Values are shown as the endpoint returned them. Method: https://radar.anlyon.com/methodology#remote-posture

| Fact | Observed |
| --- | --- |
| Reachable | `yes` |
| HTTP status of an anonymous initialize | `200` |
| tools/list served without credentials | `yes` |
| Tools that change things, listed without credentials | `30` |
| Protocol version | `2025-06-18` |
| WWW-Authenticate on 401 | `not observed` |
| resource_metadata in WWW-Authenticate | `none` |
| RFC 9728 protected resource metadata | `https://seoscoreapi.com/.well-known/oauth-protected-resource/mcp` |
| Authorization servers | `https://seoscoreapi.com` |
| Legacy HTTP+SSE transport | `no` |
| Dynamic client registration | `yes` |
| Client ID metadata documents | `yes` |

- **high** Tools that change things, listed to anyone ([unauth-remote-writes](https://radar.anlyon.com/methodology#unauth-remote-writes))

```untrusted-server-text
The endpoint served tools/list without credentials, and the list has 30 tools that change things (4 destructive, 2 irreversible, 24 write). Whether a call also works without credentials was not tested: tools are never called.

https://seoscoreapi.com/mcp: tools/list returned 45 tools to an anonymous client, including remove_monitor (destructive), remove_geo_monitor (destructive), update_citation_tracker (destructive), delete_citation_tracker (destructive), citation_topups (irreversible) and 25 more
```
