# OpenAEO (npm · openaeo-audit)

Asks 8 AI crawlers for your page, checks who really answered, and writes the fixes. MIT.

- Trust score: 64/100 (medium)
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-08-07

## Components

- npm · `openaeo-audit`: 64/100 (this document), [markdown](https://verifymcp.io/servers/dev-openaeo-audit/openaeo-audit.md), [page](https://verifymcp.io/servers/dev-openaeo-audit/openaeo-audit)

## Channel facts

- Registry: `npm`
- Package: `openaeo-audit`
- Version: `1.6.2`
- Transport: `stdio`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-08-07.

- **Supply Chain Security**: 100/100
  - No malware found by supply-chain analysis.
  - No known CVEs affecting this package version or its production dependencies.
  - No install/post-install scripts declared.
  - No production dependencies, so there is no dependency health to assess.
- **Provenance & Transparency**: 45/100
  - Source repository is publicly reachable at the declared URL.
  - Provenance check failed: no build-provenance attestation is published.
  - Clear OSI-approved license (MIT).
  - Actively maintained (last published 0 days ago).
  - Disclosure check failed: no security disclosure policy was found in the source repository.
- **Schema Quality & AI Usability**: 64/100
  - AI-judged instruction clarity (good).
  - Context-footprint check failed: tool/resource definitions use about 2195 tokens (~146/item across 15 items; 15 tools + 0 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 7/100
  - Stability observed for 2 of 30 days with no destabilising changes; credit accrues until the full window elapses.
- **Tool Coverage**: 100/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 100% of tool parameters carry a description.
- **Capabilities**: 20/100
  - Spec-recency check failed: implements MCP spec 2024-11-05; the latest is 2026-07-28.

## Install

### Claude

```bash
claude mcp add dev-openaeo-audit -- npx -y openaeo-audit
```

### Codex

```bash
codex mcp add dev-openaeo-audit -- npx -y openaeo-audit
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "dev-openaeo-audit": {
      "type": "local",
      "command": [
        "npx",
        "-y",
        "openaeo-audit"
      ],
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add dev-openaeo-audit --command npx --arg -y --arg openaeo-audit
```

### Hermes

```yaml
mcp_servers:
  dev-openaeo-audit:
    command: "npx"
    args: ["-y", "openaeo-audit"]
```

### Other

```json
{
  "mcpServers": {
    "dev-openaeo-audit": {
      "command": "npx",
      "args": [
        "-y",
        "openaeo-audit"
      ]
    }
  }
}
```

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-08-07 (score 64, +1)

- [security regression] Stability: 0.03 → unverified
- [functional regression] Tool coverage: 100 → unverified
- [functional regression] Capabilities: fail → unverified
- [functional] Package version: 1.6.1 → 1.6.2

### 2026-08-06 (score 63, +24)

- [functional regression] MCP protocol: unverified → fail
- [functional improvement] Tool coverage: unverified → 100
- [functional improvement] Stability: unverified → 0.03
- [functional] First check of Schema quality: fail
- [functional] First check of Schema quality: fail
- [functional] First check of Schema quality: good
- [functional] First check of Tool coverage: 100
- [functional] Package version: 1.6.0 → 1.6.1

### 2026-08-05 (score 39, 0)

- [security] Stability: Stability not yet verified: we do not have a sandbox capture of the MCP schema this version of the package serves yet.
- [functional] Tool coverage: Tool coverage not yet verified: we do not have a sandbox capture of the tool definitions this version of the package serves yet.
- [functional] Capabilities: Protocol version not yet verified: we do not have a sandbox capture of the MCP handshake this version of the package performs yet.
- [functional] Schema quality: Schema quality not yet verified: we do not have a sandbox capture of the MCP schema this version of the package serves yet.
- [functional] Package version: 1.3.0 → 1.6.0
- [functional] Package version: 1.3.0 → 1.5.0

### 2026-08-04 (score 39)

First indexed and scored.

## MCP tools (15)

### `aeo_audit` (~97 tokens)

Audit a live website for AI-search visibility (AEO/GEO). Fetches the page as an AI crawler would and returns a 0-100 score, a band, every passing/failing check, and prioritised fixes. Use this first to find out why a site isn't being cited by ChatGPT, Claude, Gemini or Perplexity.

Input parameters:

- `domain` (string, required): Domain or full URL, e.g. example.com

### `aeo_fix_files` (~100 tokens)

Generate the paste-ready files a site needs to be readable and citable by AI assistants: robots.txt (allowing the AI crawlers), llms.txt (a quotable index), Organization+WebSite JSON-LD, and a template opening paragraph. Write these into the user's project.

Input parameters:

- `brand` (string): Business/brand name (optional; defaults from the domain)
- `domain` (string, required): Domain, e.g. example.com

### `aeo_fix_html` (~126 tokens)

Take a page's HTML and return it with the missing structural pieces injected into <head>: <title>, <meta name=description>, and Organization+WebSite JSON-LD. Only ADDS what's absent, never rewrites or deletes existing content. Returns the fixed HTML plus a before/after readiness score. Use this on a local file, then write the result back.

Input parameters:

- `brand` (string): Business/brand name (optional)
- `domain` (string): Domain, e.g. example.com
- `html` (string, required): The full HTML source of the page

### `aeo_fix_my_site` (~146 tokens)

ONE-SHOT for people who don't have a codebase. Give it a domain and it audits the site, detects the platform (Squarespace, Wix, Webflow, WordPress, Shopify, Framer, Ghost, Carrd or custom), generates the exact files/snippets, and returns click-by-click instructions for where to paste each one in THAT platform's admin, including an honest note when a platform can't do something. Use this whenever the user isn't a developer or their site is on a website builder.

Input parameters:

- `brand` (string): Business name (optional; defaults from the domain)
- `domain` (string, required): Domain or full URL, e.g. example.com

### `aeo_packet` (~123 tokens)

Generate an engineering packet for a domain (the open-source version): an ordered list of tickets, each with the audit evidence that justifies it, where the change goes for that stack, implementation steps, acceptance criteria, and an `agentPrompt` you can execute directly. Use this when the user wants you to actually FIX their site rather than just audit it, work the tickets in order, P1 first, showing the user each change before you make it.

Input parameters:

- `brand` (string): Business name (optional)
- `domain` (string, required): Domain or full URL

### `aeo_check_html` (~73 tokens)

Score a page's HTML against the checks a single file can satisfy (title, meta description, Organization/WebSite/FAQPage JSON-LD, a number in the first 100 words) without making changes. Use to verify your edits worked.

Input parameters:

- `html` (string, required): The full HTML source of the page

### `aeo_read_log` (~250 tokens)

Read a server access log and report what AI crawlers actually got: which arrived, what status they were given, which paths they reached, and whether each request really came from the operator it claimed. Understands the combined format nginx and Apache write by default, and JSON lines from Cloudflare, Fastly, Vercel and Netlify. This is the tool that settles what an audit can only indicate: a fetch sent from this machine wearing a crawler's name is an indication, and the log is the evidence. Where the two disagree, the log is right. Runs entirely locally. The log is parsed in this process and never sent anywhere; the only network call is fetching the operators' public address ranges.

Input parameters:

- `bots` (array): Optional: the bots array from a previous aeo_audit call, to reconcile what we measured against what actually happened.
- `domain` (string): Optional: the site this log belongs to. With an API key this adds your authenticity rate to the corpus and returns how you compare. Only the rate and the request count are sent; the log is not.
- `log` (string, required): The access log text. Paste it or read the file first.

### `aeo_verify_crawler` (~139 tokens)

Check whether an IP address really belongs to the crawler it claims to be, against the ranges the operator publishes. OpenAI, Anthropic, Google, Microsoft, Perplexity and Apple all publish these. Answers verified, impostor, or unverifiable, and never says impostor without positive evidence: if the operator publishes nothing, or the feed fails to load, the answer is unverifiable. Use when a log line looks suspicious or before allow-listing anything by name.

Input parameters:

- `bot` (string, required): The crawler it claims to be, e.g. GPTBot
- `ip` (string, required): The client IP address from your log

### `aeo_authenticity` (~153 tokens)

How much of the AI crawler traffic a site thinks it is getting is actually real, and how that compares to every other site we have read a log for. Every owner assumes their answer is 100%. The comparison is the part no local tool can produce: it needs other people's logs. Call with no arguments to read the current median, or pass a rate from aeo_read_log to add your own reading and get back where you stand.

Input parameters:

- `claimed` (number): Optional: how many requests that rate was computed over
- `domain` (string): The site the reading is for
- `rate` (number): Optional: the authenticity rate from aeo_read_log, 0 to 100

### `aeo_alerts` (~87 tokens)

What changed on this site since the last check: a crawler that could reach it last week and cannot now, a check that was passing and is not, citations lost or gained. A crawler going dark usually means nobody touched the site and a CDN moved a default underneath it, which is the one problem an owner cannot notice on their own. Worth calling on a schedule rather than once.

### `aeo_crawler_intel` (~118 tokens)

Has this site's CDN started refusing AI crawlers? Reads OpenAEO's corpus of per-crawler reachability observations to show what each crawler is served today, what changed since the last observation, and how often each has been blocked. Answers what robots.txt cannot: a CDN can refuse ClaudeBot while robots.txt allows it, and a provider changing a default is invisible to the site owner. Needs OPENAEO_API_KEY (Solo and above).

Input parameters:

- `domain` (string, required): Domain to look up, e.g. example.com

### `aeo_citations` (~92 tokens)

Run a citation test: ask the assistants a buyer question and record whether this business is named. Runs on openaeo.dev because it needs provider keys for six surfaces. Needs OPENAEO_API_KEY (Solo and above).

Input parameters:

- `domain` (string, required): The domain to look for in the answers
- `query` (string, required): The buyer question to ask, e.g. 'best film presets for portraits'

### `aeo_history` (~42 tokens)

Score history for a tracked site: how the AEO score has moved over time, and which checks changed. Needs OPENAEO_API_KEY (Solo and above).

### `aeo_drift` (~79 tokens)

What regressed. Lists checks that were passing and are not any more, and crawlers that were reachable and are not any more, so a deploy that broke AI visibility is caught before anyone notices the traffic. Needs OPENAEO_API_KEY (Solo and above).

Input parameters:

- `domain` (string): Domain to check, defaults to your tracked site

### `aeo_competitors` (~42 tokens)

Competitor share of voice: which domains the assistants name instead of you, across your tracked prompts. Needs OPENAEO_API_KEY (Business and above).

## Diagnostics

Captured diagnostic sections: Provenance, Dependencies. The full working is on the page: https://verifymcp.io/servers/dev-openaeo-audit/openaeo-audit#diagnostics

## Score history

- 2026-08-07: 64
- 2026-08-06: 63
- 2026-08-05: 39
- 2026-08-04: 39

## Links

- npm package: https://www.npmjs.com/package/openaeo-audit
- Socket report: https://socket.dev/npm/package/openaeo-audit
- Repository: https://github.com/charlacsina/openaeo-audit
- Website: https://openaeo.dev/mcp
- Changelog RSS feed: https://verifymcp.io/servers/dev-openaeo-audit/openaeo-audit.xml
- Changelog JSON feed: https://verifymcp.io/servers/dev-openaeo-audit/openaeo-audit.json
- HTML version of this page: https://verifymcp.io/servers/dev-openaeo-audit/openaeo-audit