> ## Documentation Index
> Fetch the complete documentation index at: https://nevermined.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# query_docs_filesystem_nevermined_documentation

> Docs MCP tool: run a read-only shell (rg, cat, tree, jq, …) over a virtual filesystem of the Nevermined docs pages and OpenAPI specs.

Run a **read-only, shell-like** query against a virtualized, in-memory filesystem rooted at `/` that contains **only** the Nevermined documentation pages and OpenAPI specs. This is not a shell on any real machine — nothing runs on your computer, the server host, or any network; the filesystem is a sandbox backed by documentation chunks. This is how you read a page in full: there is no separate "get page" tool. See the [Docs MCP overview](/docs/mcp/docs-mcp/overview) for how to connect.

**Charges:** nothing — public documentation only.

## Arguments

| Argument | Required | Type | Description |
| - | - | - | - |
| `command` | yes | string | A shell command to run against the virtual filesystem (e.g. `rg -il "keyword" /`, `tree / -L 2`, `head -80 /path/file.mdx`). |

## How it works

* **Read a page:** a page at URL path `/some/page` lives at `/some/page.mdx`. Read it with `head -200 /some/page.mdx` or `cat /some/page.mdx`.
* **Discover paths — never guess them:** use `tree / -L 2`, `ls /`, or [`search_nevermined_documentation`](/docs/mcp/docs-mcp/search-documentation). A path that doesn't exist only means the guess was wrong, not that the topic is undocumented — use `rg -il "keyword" /` to find it.
* **OpenAPI specs** are mounted at `/openapi/api-reference/openapi.json`, `/openapi/api-reference/oauth-openapi.json`, and `/openapi/api-reference/organizations-openapi.json`. Example: `cat /openapi/api-reference/openapi.json | jq '.paths | keys'`.

**Supported commands:** `rg` (ripgrep), `grep`, `find`, `tree`, `ls`, `cat`, `head`, `tail`, `stat`, `wc`, `sort`, `uniq`, `cut`, `sed`, `awk`, `jq`, plus basic text utilities. No writes, no network, no process control. Run `--help` on any command for usage.

<Warning>
  **Each call is stateless.** The working directory always resets to `/` and no variables, aliases, or history carry over between calls. To operate in a subdirectory, chain commands in one call with `&&` or pass absolute paths (e.g. `cd /some-dir && ls`, or `ls /some-dir`). Output is truncated to **30 KB** per call — prefer targeted `rg -C 3 "pattern" /path/file.mdx` or `head -N` over a broad `cat` on large files, and batch multiple file reads into one `head`/`cat`.
</Warning>

## Returns

The command's stdout, truncated to 30 KB. When referencing pages back to the user, convert filesystem paths to URL paths by removing the `.mdx` extension (`/some/page.mdx` → `/some/page`).

## Try it

**Ask your assistant:**

> Show the top-level structure of the Nevermined documentation.

**Call it directly** (the docs server replies over SSE — `event: message` / `data: …`):

```bash theme={null}
curl -X POST https://nevermined.ai/docs/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"query_docs_filesystem_nevermined_documentation","arguments":{"command":"tree / -L 2"}}}'
```


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.