# `ai.surveygent/surveygent` 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/ai.surveygent/surveygent
- JSON: https://radar.anlyon.com/api/v1/servers/ai.surveygent/surveygent
- Version: 1.2.0 (active)
- Publisher: ai.surveygent (domain proven to the MCP Registry)
- Runs: remote (streamable-http)
- Sign-in to list tools: none
- Repository: https://github.com/Surveygent/surveygent-core
- Metadata from the official MCP Registry, synced 2026-11-19T22:17:38.495Z

Title and description, as the publisher wrote them:

```untrusted-server-text
Surveygent
Interview and optional work receipts. No credentials or system access.
```

Scanned 2026-11-19T21:18:12.694Z over streamable-http.

**5 tools: 2 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 | 2 | 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.

| 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__surveygent__log_action` `mcp__surveygent__verify_claim` |
| outbound | Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL. | `mcp__surveygent__log_action` |

## Findings

- **info** The word "credentials" in `get_survey` ([credentials-mention](https://radar.anlyon.com/methodology#credentials-mention))

```untrusted-server-text
The description mentions "credentials".

description: "...urvey short-v1 for the short survey. No credentials are requested."
```

- **info** The word "credentials" in `submit_answers` ([credentials-mention](https://radar.anlyon.com/methodology#credentials-mention))

```untrusted-server-text
The description mentions "credentials".

description: "...t return other respondents' answers. No credentials are requested."
```

- **info** The word "credentials" in `submit_answers` ([credentials-mention](https://radar.anlyon.com/methodology#credentials-mention))

```untrusted-server-text
The input schema at properties.respondent_key.description mentions "credential".

inputSchema.properties.respondent_key.description: "...onymous key. Not a name, URL, email, or credential."
```

- **info** The word "credentials" in `log_action` ([credentials-mention](https://radar.anlyon.com/methodology#credentials-mention))

```untrusted-server-text
The description mentions "credential".

description: "...blic https URL. A field that contains a credential is rejected. No network address is stor..."
```

- **info** The word "credentials" in `log_action` ([credentials-mention](https://radar.anlyon.com/methodology#credentials-mention))

```untrusted-server-text
The input schema at properties.agent_key.description mentions "credential".

inputSchema.properties.agent_key.description: "Pseudonymous key for this log. Not a credential."
```


## Changes things (2)

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

### `log_action`

- The name "log_action" says "log": a word for reading.
- The description's first verb is "store": a word for a change.
- Tag untrusted-input: The input takes a URL (input.target_url), and the tool reads it: it reads web pages, which anyone can write.
- Tag outbound: The input takes a URL (input.target_url): a request to a URL can carry data to whoever runs it.
- Findings: info credentials-mention, info credentials-mention

Description, as the server wrote it:

```untrusted-server-text
Stores one declared action for a pseudonymous agent_key and returns a receipt. The receipt has an id, the server timestamp, and a hash chained to that agent's previous receipt, plus source, as_of, ttl, version, and label declared. The same idempotency_key for that agent_key is HTTP 409 and includes the original receipt id. target_url is a public https URL. A field that contains a credential is rejected. No network address is stored.
```

### `submit_answers`

- The name "submit_answers" says "submit": a word for a change.
- Findings: info credentials-mention, info credentials-mention

Description, as the server wrote it:

```untrusted-server-text
Records this agent's answers. q1–q5 are the common questions. Later questions depend on the answer to q5. The short survey sends role, hardest_part, one_change, anything_else, and optional tools_valued and can_buy, with survey_version short-v1. unknown and skip are valid. The tool does not return other respondents' answers. No credentials are requested.
```

## Only reads (3)

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

### `get_receipts`

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

Description, as the server wrote it:

```untrusted-server-text
Returns the receipt log stored for one pseudonymous agent_key. Each receipt includes its chain hash and the fields source, as_of, ttl, version, and label. The log is keyed only by that agent_key. No network address is included.
```

### `get_survey`

- The name "get_survey" says "get": a word for reading.
- Findings: info credentials-mention

Description, as the server wrote it:

```untrusted-server-text
Return the Surveygent interview questions. Pass survey short-v1 for the short survey. No credentials are requested.
```

### `verify_claim`

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

Description, as the server wrote it:

```untrusted-server-text
Reads one stored receipt and GETs its public target_url. The verdict is verified, not_found, mismatch, or unverifiable, with source, as_of, ttl, version, and label tested or unknown. verified requires a full-body sha256, an exact text/plain body of 16 to 200 characters, or an exact JSON scalar at a dotted path. HTTP 404 and 410 are not_found. A finished comparison that fails is mismatch. Private, loopback, link-local, and metadata addresses, including after DNS and redirects, plus timeouts, oversized bodies, and any other uncertainty, are unverifiable.
```

## Claude Code permissions

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

```json
{
  "permissions": {
    "allow": [
      "mcp__surveygent__get_survey",
      "mcp__surveygent__verify_claim",
      "mcp__surveygent__get_receipts"
    ],
    "ask": [
      "mcp__surveygent__submit_answers",
      "mcp__surveygent__log_action"
    ],
    "deny": []
  }
}
```

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

- `mcp__surveygent__log_action`: untrusted-input, outbound
- `mcp__surveygent__verify_claim`: untrusted-input

## Remote posture

What an anonymous client saw at the remote endpoint, last 2026-11-17T15:18:30.218Z, unchanged since 2026-11-17T15:18:30.218Z. 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 | `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://radar.anlyon.com/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://api.sunlyght.com/mcp: tools/list returned 2 tools to an anonymous client, including submit_answers (write)
```
