# `garden.lexicon/lexicon-garden` 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/garden.lexicon/lexicon-garden
- JSON: https://radar.anlyon.com/api/v1/servers/garden.lexicon/lexicon-garden
- Version: 3.0.20 (active)
- Publisher: garden.lexicon (domain proven to the MCP Registry)
- Runs: remote (streamable-http)
- Sign-in to list tools: none
- Metadata from the official MCP Registry, synced 2026-11-20T14:17:43.201Z

Title and description, as the publisher wrote them:

```untrusted-server-text
Lexicon Garden
Resolve, search, and validate AT Protocol lexicons.
```

Scanned 2026-11-20T14:18:27.942Z over streamable-http.

**6 tools: 2 change things and 4 only read.**

| Group | Class | Tools | Claude Code |
| --- | --- | --- | --- |
| Deletes or cancels | destructive | 0 | deny |
| Can't be undone | irreversible | 0 | ask |
| Changes things | write | 2 | ask |
| Only reads | read | 4 | allow |

## Exfil surface

No private-data -> untrusted-input -> outbound path in this server alone: no tool is tagged private-data or untrusted-input or outbound.

| Tag | Meaning | Tools |
| --- | --- | --- |
| private-data | Reads files, mail, messages, databases, secrets, calendars or repositories. | none |
| untrusted-input | Brings in text someone else can write: web pages, inbound email, messages, issues and comments. | none |
| outbound | Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL. | none |

## Findings

No check matched anything in these tool definitions.


## Changes things (2)

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

### `check_compatibility`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Check compatibility between two versions of an ATProtocol lexicon schema. Detects breaking vs non-breaking changes and provides migration guidance. Supply either two schema JSON objects directly (from_schema + to_schema), or an NSID with two CIDs to compare stored versions (nsid + from_cid + to_cid).
```

### `discover_permission_sets`

- The name "discover_permission_sets" says "set": a word for a change.

Description, as the server wrote it:

```untrusted-server-text
Discover which ATProtocol permission-set lexicon schemas cover a given set of collections and/or XRPC methods. Returns matching permission sets with their OAuth scope strings (`include:<nsid>`), useful for determining what OAuth scopes an application needs.
```

## Only reads (4)

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

### `describe_lexicon`

- The name "describe_lexicon" says "describe": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
Get detailed information about an ATProtocol lexicon schema: its definitions, authority, CID and any published documentation. Use this to understand the structure and format of XRPC methods, records, and other lexicon types.
```

### `resolve_lexicon`

- The name "resolve_lexicon" says "resolve": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
Resolve an ATProtocol lexicon by NSID: find the DID named by the `_lexicon` TXT record of the NSID's authority and read that DID's com.atproto.lexicon.schema record, from Lexicon Garden's index or, when it is not indexed or refresh is set, from the network. Returns the schema, its AT-URI and CID, the authority DID and where it came from.
```

### `search_lexicons`

- The name "search_lexicons" says "search": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
Find ATProtocol lexicons by what their schemas say: descriptions, definition names, field names and known values. Use it when you don't know the NSID. mode "keyword" (default) matches words; mode "semantic" ranks by meaning, so a plain description of what you need works (e.g. 'store a user's reading list'), and it falls back to keyword when unavailable. By default only lexicons published by their NSID's DNS authority are returned. Returns NSIDs with summaries and AT-URIs; follow up with describe_lexicon.
```

### `validate_lexicon`

- The name "validate_lexicon" says "validate": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
Validate a lexicon schema and optionally validate data against it. Use this to verify schema correctness and validate records, query params, or procedure inputs against their schemas.
```

## Claude Code permissions

For `.claude/settings.json`, assuming the server is installed as `lexicon-garden`:

```json
{
  "permissions": {
    "allow": [
      "mcp__lexicon-garden__resolve_lexicon",
      "mcp__lexicon-garden__search_lexicons",
      "mcp__lexicon-garden__describe_lexicon",
      "mcp__lexicon-garden__validate_lexicon"
    ],
    "ask": [
      "mcp__lexicon-garden__check_compatibility",
      "mcp__lexicon-garden__discover_permission_sets"
    ],
    "deny": []
  }
}
```

## Remote posture

What an anonymous client saw at the remote endpoint, last 2026-11-20T18:19:32.809Z, unchanged since 2026-11-20T18:19:32.809Z. 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 | `2` |
| Protocol version | `2025-11-25` |
| WWW-Authenticate on 401 | `not observed` |
| resource_metadata in WWW-Authenticate | `none` |
| RFC 9728 protected resource metadata | `none` |
| Authorization servers | `none` |
| Legacy HTTP+SSE transport | `no` |
| Dynamic client registration | `not observed` |
| Client ID metadata documents | `not observed` |

- **warn** 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 2 tools that change things (2 write). Whether a call also works without credentials was not tested: tools are never called.

https://lexicon.garden/mcp: tools/list returned 6 tools to an anonymous client, including check_compatibility (write), discover_permission_sets (write)
```
