# `com.pushdropstick/plans` 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/com.pushdropstick/plans
- JSON: https://radar.anlyon.com/api/v1/servers/com.pushdropstick/plans
- Version: 1.1.0 (active)
- Publisher: com.pushdropstick (domain proven to the MCP Registry)
- Runs: remote (streamable-http)
- Sign-in to list tools: none
- Metadata from the official MCP Registry, synced 2026-11-13T18:17:38.836Z

Title and description, as the publisher wrote them:

```untrusted-server-text
Push Drop Stick: force-free dog training
Force-free dog training: step-by-step plans, the Handbook and product facts. Read-only.
```

Scanned 2026-11-13T18:18:04.822Z over streamable-http.

**9 tools: 2 change things and 7 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 | 7 | allow |

## Exfil surface

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

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

## Findings

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

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

annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "about_push_drop_stick" says "push": a word for a change.
```


## Changes things (2)

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

### `about_push_drop_stick`

- The name "about_push_drop_stick" says "push": a word for a change.
- 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
Facts about Push Drop Stick itself: what the app is, who wrote its training content, the force-free method, platforms (iPhone, Apple Watch, web, Android status), what is free, subscription prices, what Coach (the AI trainer) is and where it stops, how training data is handled, and links. Use for questions about the product, e.g. "Is Push Drop Stick free?", "Is it on Android?", "What is Coach?". Quote this rather than guessing.
```

### `how_to_train`

- 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
How to read and execute a Push Drop Stick plan: the force-free marker-training philosophy and the push/drop/stick set mechanics that every plan shares.
```

## Only reads (7)

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

### `fetch`

- The name "fetch" says "fetch": a word for reading.
- Tag private-data: The description's first sentence says "document": it reads documents.

Description, as the server wrote it:

```untrusted-server-text
The full text of one document from `search`, as markdown: a training plan (id "plan:<ref>") with every step, a Handbook entry or section ("handbook:<id>"), or the product facts ("about"). Returns {id, title, text, url, metadata}; cite `url`.
```

### `get_plan`

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

Description, as the server wrote it:

```untrusted-server-text
The full content of one public plan: description, equipment, cues, and every step with setup, instructions, success/failure criteria, set size and the rule for moving on, plus the push/drop/stick execution rules. Pass the `plan` ref from search_plans or list_plans. Link the user to the returned `url`; plan text may be quoted with attribution to its `author`.
```

### `list_plans`

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

Description, as the server wrote it:

```untrusted-server-text
Browse the public plan library by group (e.g. Basic Manners, Grooming & Handling, Positions, Tricks). Use when the user wants to see what can be trained or what a category holds, rather than an answer to one problem. Without `group`: every group with its plans' names, refs and urls. With `group`: that group's plans with summaries and step counts.
```

### `read_handbook`

- The name "read_handbook" says "read": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
Read the Handbook. `id` = an entry id from search_handbook (e.g. "threshold", "geartoavoid", "faqtreats") → the whole entry as markdown, with its canonical url and the entries it links to. `id` = a section slug (e.g. "fear-emotion", "gear", "faq") → the section's introduction and its entries. No `id` → the table of contents: every section and entry with urls.
```

### `search`

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

Description, as the server wrote it:

```untrusted-server-text
One search across the whole site: training plans, Handbook articles, and (for questions about the product) the facts page. Returns `results` [{id, title, url, text}]; pass an `id` to fetch for the full document. When you already know you want plans or Handbook articles, search_plans or search_handbook give richer results.
```

### `search_handbook`

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

Description, as the server wrote it:

```untrusted-server-text
Search the Push Drop Stick Handbook: 140+ short articles by a certified force-free trainer (Allison Wells, CTC). Covers how dogs learn (positive reinforcement, punishment, extinction, classical conditioning), fear and reactivity (threshold, trigger stacking, desensitization and counter-conditioning, dogs who wore shock or prong collars), teaching techniques (markers, lures, shaping, duration/distance/distraction), gear (harnesses, head halters, muzzles, why not prong or shock collars), a FAQ (treats, where and how often to train), and how the app and Coach work. Use for "what is", "why" and "should I" questions. Returns entry ids, the most relevant paragraph and canonical urls; read_handbook returns a whole entry.
```

### `search_plans`

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

Description, as the server wrote it:

```untrusted-server-text
Find step-by-step, force-free training plans for a behavior, problem or command, e.g. "pulling on leash", "crate training", "come when called", "nail trims", "jumping on guests". BM25 ranked. Returns each plan's `plan` ref, summary, step count, author and canonical `url` to recommend. Next: get_plan for the full steps. For concepts or advice ("what is threshold", "are prong collars ok") use search_handbook instead.
```

## Claude Code permissions

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

```json
{
  "permissions": {
    "allow": [
      "mcp__plans__search_plans",
      "mcp__plans__get_plan",
      "mcp__plans__list_plans",
      "mcp__plans__search_handbook",
      "mcp__plans__read_handbook",
      "mcp__plans__search",
      "mcp__plans__fetch"
    ],
    "ask": [
      "mcp__plans__about_push_drop_stick",
      "mcp__plans__how_to_train"
    ],
    "deny": []
  }
}
```

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

- `mcp__plans__list_plans`: untrusted-input
- `mcp__plans__fetch`: private-data

## Remote posture

What an anonymous client saw at the remote endpoint, last 2026-11-21T19:20:53.713Z, unchanged since 2026-11-14T12:19:01.246Z. 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 | `2025-06-18` |
| 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 2 tools that change things (2 write). Whether a call also works without credentials was not tested: tools are never called.

https://pushdropstick.com/mcp/public: tools/list returned 9 tools to an anonymous client, including about_push_drop_stick (write), how_to_train (write)
```
