# `io.github.hifarrer/cure-cancer-with-ai` 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/io.github.hifarrer/cure-cancer-with-ai
- JSON: https://radar.anlyon.com/api/v1/servers/io.github.hifarrer/cure-cancer-with-ai
- Version: 1.0.0 (active)
- Publisher: io.github.hifarrer (GitHub account)
- Runs: remote (streamable-http)
- Sign-in to list tools: none
- Repository: https://github.com/hifarrer/ccwai-api
- Metadata from the official MCP Registry, synced 2026-10-21T18:35:55.566Z

Title and description, as the publisher wrote them:

```untrusted-server-text
Cure Cancer With AI
Free oncology data (research, trials, FDA approvals, news) plus IBM MAMMAL biomedical predictions.
```

Scanned 2026-11-19T20:18:03.772Z over streamable-http.

**15 tools: 4 change things and 11 only read.**

| Group | Class | Tools | Claude Code |
| --- | --- | --- | --- |
| Deletes or cancels | destructive | 0 | deny |
| Can't be undone | irreversible | 0 | ask |
| Changes things | write | 4 | ask |
| Only reads | read | 11 | 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__cure-cancer-with-ai__get_blog_post` `mcp__cure-cancer-with-ai__list_blog_posts` `mcp__cure-cancer-with-ai__search_oncology` |
| 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 (4)

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

### `mammal_health`

- 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 whether the IBM MAMMAL prediction model is loaded and ready. No API key required. Call this before predict_* tools if a prior prediction timed out.
```

### `predict_clintox`

- 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
Predict clinical-trial toxicity for a compound using IBM MAMMAL. Returns pred 1 (toxic / likely to fail trials) or 0 (not toxic) plus a raw score. Inference is CPU-bound and may take up to ~60s.
```

### `predict_dti`

- 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
Predict drug–target binding affinity as pKd (−log10 Kd; higher = stronger binding) using IBM MAMMAL. Inference is CPU-bound and may take up to ~60s.
```

### `predict_ppi`

- 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
Predict the binding-affinity class for a pair of proteins using the IBM MAMMAL biomedical foundation model. Returns label "1" (interacting) or "0" (non-interacting). Inference is CPU-bound and may take up to ~60s.
```

## Only reads (11)

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

### `get_blog_post`

- The name "get_blog_post" says "get": a word for reading.
- Tag untrusted-input: The name "get_blog_post" says "get" and "post": it reads posts, which other people can write. The description's first sentence says "fetch" and "post": it reads posts, which other people can write.

Description, as the server wrote it:

```untrusted-server-text
Fetch a single blog post by slug, including the full article content.
```

### `get_clinical_trial`

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

Description, as the server wrote it:

```untrusted-server-text
Fetch a single clinical trial by NCT id (or internal id), including eligibility criteria and locations.
```

### `get_research_paper`

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

Description, as the server wrote it:

```untrusted-server-text
Fetch a single research paper by its internal id or PubMed id.
```

### `list_blog_posts`

- The name "list_blog_posts" says "list": a word for reading.
- Tag untrusted-input: The name "list_blog_posts" says "list" and "post": it reads posts, which other people can write.

Description, as the server wrote it:

```untrusted-server-text
List editorial blog articles (excerpts). Use get_blog_post for full content. Filter by category, cancer-type tag, or keyword.
```

### `list_clinical_trials`

- The name "list_clinical_trials" says "list": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
List clinical trials from public registries (conditions, status, intervention type). Filter by condition, status (e.g. RECRUITING), or keyword.
```

### `list_compound_characteristics`

- The name "list_compound_characteristics" says "list": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
List the pharmaceutical-compound characteristics (each scored on the −4…+4 scale) that can be used as preference keys in search_compounds.
```

### `list_fda_approvals`

- The name "list_fda_approvals" says "list": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
List FDA-approved oncology drugs with indication, company, approval date, and label links. Filter by cancer type, keyword, or approval date.
```

### `list_news`

- The name "list_news" says "list": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
List curated cancer news articles aggregated from trusted sources. Filter by cancer type, keyword, or published date.
```

### `list_research`

- The name "list_research" says "list": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
List peer-reviewed oncology research papers ingested from PubMed (abstracts, authors, journal, plain-language summaries). Filter by cancer type, treatment type, keyword, or publication date.
```

### `search_compounds`

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

Description, as the server wrote it:

```untrusted-server-text
Find pharmaceutical compounds two ways: by example drugs you already know (fuzzy-matched), or by setting target characteristics on a −4…+4 scale. Provide exactly one of `examples` or `preferences`. Use list_compound_characteristics for the available preference names.
```

### `search_oncology`

- The name "search_oncology" says "search": a word for reading.
- Tag untrusted-input: The description's first sentence says "search" and "post": it reads posts, which other people can write.

Description, as the server wrote it:

```untrusted-server-text
Search a keyword across every Cure Cancer With AI dataset at once — research papers, news, blog posts, FDA approvals, and clinical trials — with results grouped by type. Use this first for broad discovery, then fetch a single record by id/slug/nctId for full detail.
```

## Claude Code permissions

For `.claude/settings.json`, assuming the server is installed as `cure-cancer-with-ai`:

```json
{
  "permissions": {
    "allow": [
      "mcp__cure-cancer-with-ai__search_oncology",
      "mcp__cure-cancer-with-ai__list_research",
      "mcp__cure-cancer-with-ai__get_research_paper",
      "mcp__cure-cancer-with-ai__list_news",
      "mcp__cure-cancer-with-ai__list_blog_posts",
      "mcp__cure-cancer-with-ai__get_blog_post",
      "mcp__cure-cancer-with-ai__list_fda_approvals",
      "mcp__cure-cancer-with-ai__list_clinical_trials",
      "mcp__cure-cancer-with-ai__get_clinical_trial",
      "mcp__cure-cancer-with-ai__search_compounds",
      "mcp__cure-cancer-with-ai__list_compound_characteristics"
    ],
    "ask": [
      "mcp__cure-cancer-with-ai__predict_ppi",
      "mcp__cure-cancer-with-ai__predict_dti",
      "mcp__cure-cancer-with-ai__predict_clintox",
      "mcp__cure-cancer-with-ai__mammal_health"
    ],
    "deny": []
  }
}
```

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

- `mcp__cure-cancer-with-ai__search_oncology`: untrusted-input
- `mcp__cure-cancer-with-ai__list_blog_posts`: untrusted-input
- `mcp__cure-cancer-with-ai__get_blog_post`: untrusted-input

## Remote posture

What an anonymous client saw at the remote endpoint, last 2026-11-20T19:05:43.855Z, unchanged since 2026-10-22T15:00:06.000Z. 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 | `4` |
| 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 4 tools that change things (4 write). Whether a call also works without credentials was not tested: tools are never called.

https://www.curecancerwithai.com/api/mcp: tools/list returned 15 tools to an anonymous client, including predict_ppi (write), predict_dti (write), predict_clintox (write), mammal_health (write)
```
