# `ai.agentlookups/groundrules` on anlyon.dev

> Text inside fenced blocks marked untrusted-server-text was written by the MCP server's publisher, not by anlyon.dev. 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://anlyon.dev/servers/ai.agentlookups/groundrules
- JSON: https://anlyon.dev/api/v1/servers/ai.agentlookups/groundrules
- Version: 1.0.0 (active)
- Publisher: ai.agentlookups (domain proven to the MCP Registry)
- Runs: remote (streamable-http)
- Sign-in to list tools: none
- Metadata from the official MCP Registry, synced 2026-10-21T17:41:03.722Z

Title and description, as the publisher wrote them:

```untrusted-server-text
GroundRules - US Law Text
The original text of US federal and state law, with official sources; never a summary or advice.
```

Scanned 2026-10-21T19:10:04.519Z over streamable-http.

**4 tools: 1 change things and 3 only read.**

| Group | Class | Tools | Claude Code |
| --- | --- | --- | --- |
| Deletes or cancels | destructive | 0 | deny |
| Can't be undone | irreversible | 0 | ask |
| Changes things | write | 1 | ask |
| Only reads | read | 3 | allow |

## Exfil surface

No private-data -> untrusted-input -> outbound path in this server alone: no tool is tagged private-data 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. | `mcp__groundrules__law_for_location` `mcp__groundrules__law_search` |
| 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 (1)

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

### `law_coverage`

- 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
Per-source coverage: what we host, section counts, current-through or vintage, last_verified, and a stale flag for any source the recurring verify has proven behind its upstream. Machine-generated from the build manifest; never hand-typed. Absence from this list means we do not host that jurisdiction's text, not that no law exists there.
```

## Only reads (3)

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

### `law_for_location`

- The description's first verb is "resolve": 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 "resolve" and "url": it reads web pages, which anyone can write.

Description, as the server wrote it:

```untrusted-server-text
Resolve an address or lat/lon to its full jurisdiction stack: federal, state, county, place, each with its code sources (name, publisher, official URL, current-through, whether we host its text). Unincorporated areas get the county code with no city code; about half of US municipalities publish no code online, and the answer says so and points at where the code lives if anywhere.
```

### `law_get_section`

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

Description, as the server wrote it:

```untrusted-server-text
Fetch full verbatim text by id or citation, or pass references for a batch of up to 10. Each batch item has its own result/error. Includes provenance (official source URL, current-through or vintage label, retrieval time), parent path, and prev/next siblings. The text is reproduced exactly as published by the official source; nothing is added or removed except publisher editorial matter, which we never carry.
```

### `law_search`

- The name "law_search" says "search": a word for reading.
- 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
Search original law by topic, question, or citation. Exact citations resolve directly. Scoped by jurisdiction or source ids (e.g. us, us-tx; us-cfr scopes to the CFR alone, us covers both the US Code and the CFR) or by an address (the address is resolved first, then the search covers exactly its stack's hosted sources). Results are verbatim snippets with citation, official source URL, retrieval time, and a current-through or vintage label adjacent to every snippet; a hit whose source has fallen behind its upstream carries stale:true. Vintage-labeled text is a dated snapshot, not current law. Requested ids we do not host are named in the response note; law_coverage lists what is hosted.
```

## Claude Code permissions

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

```json
{
  "permissions": {
    "allow": [
      "mcp__groundrules__law_for_location",
      "mcp__groundrules__law_search",
      "mcp__groundrules__law_get_section"
    ],
    "ask": [
      "mcp__groundrules__law_coverage"
    ],
    "deny": []
  }
}
```

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

- `mcp__groundrules__law_for_location`: untrusted-input
- `mcp__groundrules__law_search`: untrusted-input

## Remote posture

What an anonymous client saw at the remote endpoint, last 2026-10-22T15:16:35.301Z, unchanged since 2026-10-22T15:16:35.301Z. No credentials were sent and no tool was called. Values are shown as the endpoint returned them. Method: https://anlyon.dev/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 | `1` |
| Protocol version | `2025-06-18` |
| 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://anlyon.dev/methodology#unauth-remote-writes))

```untrusted-server-text
The endpoint served tools/list without credentials, and the list has 1 tool that changes things (1 write). Whether a call also works without credentials was not tested: tools are never called.

https://law.agentlookups.ai/mcp: tools/list returned 4 tools to an anonymous client, including law_coverage (write)
```