# readOnlyHint, destructiveHint and openWorldHint are explicit booleans

**OpenAI requires readOnlyHint, destructiveHint and openWorldHint as explicit true or false booleans in every tool’s annotations, even where the MCP specification has a default; the submission form will not continue until they are set.**

Check `openai-hints-explicit` · OpenAI plugin directory · default severity: blocking · rubric 2.1.0. Last verified against OpenAI plugin guidelines, section 6.3 on 2026-11-18.

## Error message you’ll see

```text
Each MCP tool must set readOnlyHint, openWorldHint, and destructiveHint to true or false.
```

ChatGPT app submission form (blocks Continue until every tool has all three). Where this was reported: https://community.openai.com/t/mcp-submission-blocked-each-mcp-tool-must-set-readonlyhint-openworldhint-destructivehint-despite-correct-server-config/1379193

## What fails

OpenAI requires all three hints on every tool, even where the specification has a default.

## The rule

> Set readOnlyHint, destructiveHint, and openWorldHint to the appropriate explicit boolean values (true or false) in each tool’s annotations object.

Source: [OpenAI plugin guidelines, section 6.3](https://developers.openai.com/plugins/plugin-guidelines)

## How to fix it

Add every missing hint with an explicit true or false.

1. On every tool, set all three hints explicitly.
2. readOnlyHint: true only for retrieval, computation or previews; destructiveHint: true for deletes, overwrites, cancellations, revocations and irreversible sends.
3. openWorldHint: true when the tool reaches outside your own system (public web, third-party services, other users), false for a closed system.

Before:

```json
"annotations": { "readOnlyHint": true }
```

After:

```json
"annotations": { "readOnlyHint": true, "destructiveHint": false, "openWorldHint": false }
```

### TypeScript SDK

```ts
import { McpServer } from '@modelcontextprotocol/server';
import * as z from 'zod';

const server = new McpServer({ name: 'issues', version: '1.0.0' });

server.registerTool(
  'search_issues',
  {
    title: 'Search issues',
    description: 'Search issues in the user’s workspace by text. Use when the user describes an issue but has no ID.',
    inputSchema: z.strictObject({
      query: z.string().describe('Words to search for'),
    }),
    annotations: { readOnlyHint: true, destructiveHint: false, openWorldHint: false },
  },
  async ({ query }) => ({ content: [{ type: 'text', text: 'Done' }] }),
);
```

### Python SDK

```python
from typing import Annotated

from pydantic import Field
from mcp.server.mcpserver import MCPServer
from mcp.types import ToolAnnotations

server = MCPServer("issues")

@server.tool(
    title="Search issues",
    annotations=ToolAnnotations(readOnlyHint=True, destructiveHint=False, openWorldHint=False),
)
def search_issues(
    query: Annotated[str, Field(description="Words to search for")],
) -> str:
    """Search issues in the user’s workspace by text. Use when the user describes an issue but has no ID."""
    return "Done"
```

## Questions

### What does openWorldHint mean?

Whether the tool interacts with an open world of external entities (the web, third-party services, other people’s data) rather than a closed domain the server controls. A web search tool is open-world; a lookup in your own database is not.

### I set all three and the form still complains. Why?

Check that the values are JSON booleans, not strings, on every tool including later pages of tools/list, and that the submitted URL serves the updated server. A platform-side bug causing this was reported and fixed in the OpenAI developer forum.

### Does FastMCP set these?

Only if you pass them: @mcp.tool(annotations={"readOnlyHint": True, "destructiveHint": False, "openWorldHint": False}).

## Related checks

- [Every tool carries the applicable hint](https://isokron.ai/mcp-check/checks/anthropic-hint)
- [Tool fields have the types the schema defines](https://isokron.ai/mcp-check/checks/spec-field-types)
- [Read-only tools set destructiveHint: false](https://isokron.ai/mcp-check/checks/openai-readonly-destructive)
- [MCP tool annotations](https://isokron.ai/mcp-check/checks/annotations)

Counts toward: OpenAI plugin directory.

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

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