# No instructions fetched from outside sources

**A description must not tell the model to fetch or follow instructions from a URL or other outside source, because whoever controls that source can change the model’s behavior after review; put what the model needs in the description.**

Check `sec-external-instructions` · Poisoning and hidden text · default severity: blocking · rubric 2.1.0. Last verified against Anthropic connector pre-submission checklist on 2026-11-18.

## What fails

The text tells the model to fetch or follow instructions or rules from a URL or another outside source, which lets their author change the model's behavior after review.

## The rule

> Direct Claude to pull behavioral instructions from external sources

Source: [Anthropic connector pre-submission checklist](https://claude.com/docs/connectors/building/review-criteria#avoid-prompt-injection-patterns)

## How to fix it

Put everything the model needs in the description itself.

1. Remove "read the instructions at ..." and "follow the rules from ..." wording.
2. Copy any needed guidance into the description itself, briefly.
3. Links to API reference docs for custom query tools are fine; links to behavior rules are not.

A CI test that fails when the pattern comes back:

### TypeScript (node:test)

```js
// tools-list.test.mjs: run `node --test` after saving the answer with
// npx @modelcontextprotocol/inspector --cli <server> --method tools/list > tools.json
import assert from 'node:assert/strict';
import { readFileSync } from 'node:fs';
import { test } from 'node:test';

const answer = JSON.parse(readFileSync('tools.json', 'utf8'));
const tools = answer.tools ?? answer.result?.tools ?? answer;

// Every string and every key in a tool, with its path.
function* strings(value, path) {
  if (typeof value === 'string') yield [path, value];
  else if (value && typeof value === 'object')
    for (const [k, v] of Object.entries(value)) {
      yield [`${path}.${k} (key)`, k];
      yield* strings(v, `${path}.${k}`);
    }
}

const PATTERN = /(fetch|read|load|follow|get)\b[^.]{0,40}\b(instructions|rules|guidelines|prompt)\b[^.]{0,40}https?:\/\//i;

test('no instructions loaded from URLs', () => {
  const hits = [];
  for (const tool of tools) for (const [path, text] of strings(tool, tool.name)) if (PATTERN.test(text)) hits.push(path);
  assert.deepEqual(hits, [], 'instructions pulled from an outside source can change after review');
});
```

### Python (pytest)

```python
# test_tools_list.py: run `pytest` after saving the answer with
# npx @modelcontextprotocol/inspector --cli <server> --method tools/list > tools.json
import json
import re

with open("tools.json", encoding="utf-8") as f:
    answer = json.load(f)
tools = answer.get("tools") or answer.get("result", {}).get("tools") if isinstance(answer, dict) else answer


def strings(value, path):
    """Every string and every key in a tool, with its path."""
    if isinstance(value, str):
        yield path, value
    elif isinstance(value, dict):
        for k, v in value.items():
            yield f"{path}.{k} (key)", k
            yield from strings(v, f"{path}.{k}")
    elif isinstance(value, list):
        for i, v in enumerate(value):
            yield from strings(v, f"{path}.{i}")


PATTERN = re.compile(r"(?i)(fetch|read|load|follow|get)\b[^.]{0,40}\b(instructions|rules|guidelines|prompt)\b[^.]{0,40}https?://")


def test_no_instructions_loaded_from_urls():
    hits = [path for tool in tools for path, text in strings(tool, tool["name"]) if PATTERN.search(text)]
    assert hits == [], "instructions pulled from an outside source can change after review"
```

## Questions

### How is this different from linking API docs?

A link that tells a reader which API a query tool targets is required by Anthropic for custom query tools. A link the model is told to fetch and obey is what is rejected.

### What is an MCP rug pull?

A server that changes tool behavior or descriptions after it was approved. Instructions loaded from a URL are a rug pull waiting to happen.

## Related checks

- [No wording that overrides instructions or hides actions from the user](https://isokron.ai/mcp-check/checks/sec-instruction-phrases)
- [Custom query tools name or link the target API](https://isokron.ai/mcp-check/checks/anthropic-api-docs)
- [No encoded text blocks in metadata](https://isokron.ai/mcp-check/checks/sec-encoded-blob)
- [Tool poisoning and hidden text](https://isokron.ai/mcp-check/checks/security)

Counts toward: Poisoning and hidden text, Anthropic connector directory, OpenAI plugin directory.

Rubric: https://isokron.ai/mcp-check/rubric

Check your tool list: https://isokron.ai/mcp-check
