Evil MCP Server

by promptfoo

MCP server that simulates malicious behaviors for security testing

34Node.jsstdio

context tax

pending

security

97/100

cold start

pending

freshness

Active10d ago

Our reading of Evil, from the data

Evil is a security MCP server from promptfoo.

We haven't measured its tool list yet, so its token cost is unknown.

Maintenance is active: the last public signal was 10 days ago (2026-10-08). For comparison, 67% of security servers are active or maintained.

With 34 GitHub stars it ranks #103 of 1,263 security servers we index.

It runs locally on your machine over stdio.

Install Evil MCP server

claude mcp add evil -- npx -y @promptfoo/evil-mcp-server

{
  "mcpServers": {
    "evil": {
      "command": "npx",
      "args": [
        "-y",
        "@promptfoo/evil-mcp-server"
      ]
    }
  }
}

Add to ~/.cursor/mcp.json (global) or .cursor/mcp.json (project).

{
  "mcpServers": {
    "evil": {
      "command": "npx",
      "args": [
        "-y",
        "@promptfoo/evil-mcp-server"
      ]
    }
  }
}

Settings → Developer → Edit Config (claude_desktop_config.json), then restart.

{
  "servers": {
    "evil": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@promptfoo/evil-mcp-server"
      ]
    }
  }
}

Add to .vscode/mcp.json in your workspace.

{
  "mcpServers": {
    "evil": {
      "command": "npx",
      "args": [
        "-y",
        "@promptfoo/evil-mcp-server"
      ]
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json.

Security signals

97/100 — no known issues found as of 2026-10-19, from the checks below. This is not a guarantee of safety; review what a server can do before connecting it.

Deductions

No security policy (SECURITY.md)-3

Checks

✓Repository hygiene (archived, activity, license, security policy)
○Package metadata (install scripts, provenance) — pending
○Tool descriptions (prompt-injection and hidden-text heuristics) — pending
○Source scan (secrets, CVEs, risky code) by vectra-guard — pending

Scan it yourself before installing with vectra-guard, the open-source agent security CLI. Scoring rules

Freshness

Active — last maintenance signal 10d ago. The newest of the signals below sets the band.

Last commit (default branch)
2026-10-08 · 10d ago · GitHub
Latest release
2026-10-08 · 10d ago · GitHub · 1.0.5
Package published
no data · npm/PyPI
Registry entry updated
no data · official registry

How it compares in Security

Against the most-starred security servers we index. Token cost is per model turn, measured from each server's live tool list.

Server	Stars	Tools	Tokens / turn	Freshness	Runs
Evil (this)	34	—	not measured	Active	local
Semgrep	17,123	—	not measured	Active	—
Ida Pro	13,002	—	not measured	Active	—
Ghidra	11,390	—	not measured	Active	—
Robin: AI-Powered Dark Web OSINT	7,591	—	not measured	Active	local
Claude Code Ultimate Guide	6,196	—	not measured	Active	local

All security MCP servers

FAQ

›How do I install the Evil MCP server in Claude Code?

Run: claude mcp add evil -- npx -y @promptfoo/evil-mcp-server. For Cursor, VS Code, Claude Desktop and Windsurf, use the install tabs above.

›Does Evil require an API key?

No required environment variables are declared in its published metadata.

›Can I use Evil as a remote (hosted) MCP server?

No hosted endpoint is published; it runs locally over stdio.

›Is Evil in the official MCP registry?

No. It was indexed from npm.

Sources and method

GitHub repository: fetched 2026-10-19
Package registry (npm / PyPI): listed
Security signals, computed by us: 2026-10-19
Last checked against all sources: 2026-10-19

The description, packages and links come from the maintainer's own published metadata. Token cost, security signals, freshness and the reading above are MCPSkillHub's own measurements; see the methodology. Spot an error? Request a correction or claim the listing.