Atlas / MCP servers / marimo-team / codemirror-mcp

codemirror-mcpSAFE

mcp/marimo-team/codemirror-mcp

CodeMirror extension to hook up a Model Context Provider (MCP)

Verdict
SAFE
Grade
B
Trust score
89 /100
Exposed tools
4 4r · 0w · 0d
Transport
—
License
Apache-2.0
Stars
79
01

Overview

From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.

A CodeMirror extension that implements the Model Context Protocol (MCP) for resource mentions and prompt commands.

Features

  • Resource Completion: Autocomplete for @resource mentions
  • Resource Decorations: Visual styling for @resource mentions with click handling
  • Prompt Completion: Autocomplete for /prompt commands
  • Theme Support: Customizable styling

Installation

npm install @marimo-team/codemirror-mcp @modelcontextprotocol/sdk
# or
pnpm add @marimo-team/codemirror-mcp @modelcontextprotocol/sdk

Peer Dependencies

This module requires the following peer dependencies:

  • @codemirror/view
  • @codemirror/state
  • @modelcontextprotocol/sdk

Usage

import { WebSocketClientTransport } from "@modelcontextprotocol/sdk/client/websocket.js";
import { mcpExtension, extractResources } from '@marimo-team/codemirror-mcp';
import { EditorView } from '@codemirror/view';

const transport = new WebSocketClientTransport(new URL('ws://localhost:8080'));

const view = new EditorView({
extensions: [
// ... other extensions

mcpExtension({
// Required options
transport: transport,

// Optional options
logger: console,
clientOptions: {
name: 'your-client',
version: '1.0.0'
},
onResourceClick: (resource) => {
// Open resource
// e.g. open in a tab, etc.
},
}),

// Handle submit
keymap.of([
{
key: 'Enter',
run: () => {
const resources = extractResources(view);
const formattedResources = resources
.map(
({ resource }) =>
`${resource.uri} (${resource.type}): ${resource.description || resource.name}`
)
.join('\n');
const prompt = `${view.state.doc.toString()}\n\nResources:\n${formattedResources}`;
// ... submit prompt to AI server
// const response = await g
Read from source at commit a8a9a9f0768aOBSERVED · 2026-10-12
02

Connect

Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control.

claude-code
claude mcp add codemirror-mcp -- npx -y @marimo-team/[email protected]
claude-desktop
{
  "mcpServers": {
    "codemirror-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@marimo-team/[email protected]"
      ]
    }
  }
}
03

Exposed tools (4)

4 read · 0 write · 0 destructive.

ToolRiskDescription
file.txtreadA file
prompt1readTest prompt 1
prompt2readTest prompt 2
test.txtreadA test file
04

Trust audit

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

LayerWhat it checksResult
L0Provenance & inventoryPASS
L1Static analysis of the codePASS
L2Instruction surface (what it tells the agent)PASS
L3Class-specific surfacePASS
L4Behavioural (sandbox)SKIPPED

What the source does

Filesystem
none-observed
Network
none-observed
Shell
none-observed
Dependencies
not all pinned
Secrets in source
none-found

Findings (2)

LOWInventory / provenance · inv.symlink · CWE-1104
CLAUDE.md
CLAUDE.md
Why it matters. link not followed
LOWSupply chain · supply.unpinned · CWE-829, CWE-1357
package.json
@biomejs/biome, @codemirror/lang-javascript, @codemirror/lang-markdown, @codemirror/view, @modelcontextprotocol/sdk, @types/node, @uiw/codemirror-extensions-mentions, codemirror
Why it matters. 14 dependency range(s) float
Fix. pin exact versions or ship a lockfile

Gates applied: no_behavioural_pass.

Audited 2026-10-12 · audit v0.4.1 · source sha a8a9a9f0768afull audit observations/trust-audit/mcp-server/marimo-team__codemirror-mcp.json · Report an issue / request a re-scan
05

Audit history

Every audit this server has had. A grade with a past is a grade somebody is still checking.

DateSourceVerdictGradeScoreChange
2026-10-12a8a9a9f0768aSAFEB89first audit
06

Questions

What is the @marimo-team/codemirror-mcp MCP server?

CodeMirror extension to hook up a Model Context Provider (MCP)

What tools does @marimo-team/codemirror-mcp expose?

4 in total: 4 read-only, 0 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.

Is @marimo-team/codemirror-mcp safe to connect to an agent?

The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B.

What credentials does @marimo-team/codemirror-mcp need?

No credential environment variables were found in its source, so it appears to need none.

How current is this page?

The grade is for one exact copy of the source (a8a9a9f0768a), read on 2026-10-12. The repository is watched and re-audited when it changes.

Advertisement