> For the complete documentation index, see [llms.txt](https://docs.feathery.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.feathery.io/develop/feathery-mcp/security-limits-and-troubleshooting.md).

# Security, Limits and Troubleshooting

Feathery MCP uses your Feathery API key. Treat an MCP connection like any other integration that can access your Feathery data.

### Keep Your API Key Safe

* Use a Feathery API key, not an SDK key.
* Store the key in your MCP client's secret or environment variable settings when possible.
* Do not commit API keys to shared configuration files.
* Use a test environment API key when experimenting.
* Rotate the API key if it is shared accidentally.

### Data Access

The assistant can only access the Feathery account and environment associated with the API key you provide.

Some actions also depend on your Feathery plan and enabled features. For example, AI document extraction tools require Document Intelligence.

### Actions That Change Data

Some tools can create, update, or delete Feathery data. Review these actions before approving them in your AI client.

Tools that can update data include:

* Form creation, updates, copies, and deletion
* Submission creation and updates
* User creation and deletion
* Hidden field creation, updates, and deletion
* Team member invites, role changes, and removal
* AI document extraction runs
* Submission PDF exports

Deletion and removal tools require an explicit confirmation value before they run.

### File Uploads For AI Extractions

When running an AI document extraction, you can provide files from public URLs or upload local files through your MCP client.

If a public URL does not work, try downloading the file locally and attaching it through your MCP client instead.

### Common Issues

| Issue                                                   | What to check                                                                              |
| ------------------------------------------------------- | ------------------------------------------------------------------------------------------ |
| Feathery does not appear in your AI client              | Confirm the MCP server was saved, then restart or reload the client.                       |
| Authentication fails                                    | Confirm the API key is valid and the authorization value starts with `Token` .             |
| The assistant cannot find expected forms or submissions | Confirm you connected with the API key for the correct Feathery environment.               |
| An action is unavailable                                | Confirm the action is listed in Feathery MCP - Supported Tools.                            |
| AI extraction tools fail                                | Confirm Document Intelligence is enabled for your account and that the file is accessible. |
| The assistant is rate limited                           | Ask it to retry more slowly or narrow the request.                                         |
| A delete action does not run                            | Confirm the assistant included the required confirmation value.                            |

### Frequently Asked Questions

#### Can the assistant access production data?

Yes, if you connect with a live environment API key. Use a test environment API key for testing or dry runs.

#### Can I use an SDK key?

No. Use a Feathery API key.

#### Can I limit what the assistant can access?

Use the API key for the environment you want the assistant to access. Do not connect a live API key if you only want the assistant to work with test data.

#### Why are some Feathery features missing?

Feathery MCP currently supports a focused set of forms, submissions, users, fields, team account, AI extraction, document template, and log tools. More Feathery actions may be added over time.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.feathery.io/develop/feathery-mcp/security-limits-and-troubleshooting.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.