KillerSkills
Security verdicts and a runtime guard for agent skills and MCP servers before they run
- Pricing
- Freemium
- Platforms
- CLI, macOS, Linux
- Runs
- Locally
- Version
- 0.15.0
- Listing updated
- 3 days ago
About KillerSkills
KillerSkills is a freemium MCP server for security. It runs locally on your machine and is started by your AI client.
Where most MCP servers add capability, vet-skills adds a checkpoint. Its verdicts come from KillerSkills sandbox runs, and the CLI checks a skill or MCP server before it is installed. The headline feature, vet-skills guard install, rewrites the MCP configuration of 14 agents including Claude Code, Cursor, Codex, Gemini CLI and Zed so each stdio server launches under a guard.
The guard replaces credentials in the server's environment with placeholders and injects the real value only into requests to the vendor's own hosts, so a placeholder sent elsewhere is logged as an exfiltration attempt. It records every host a server contacts, refuses versions with a Blocked verdict, incident or revocation, and applies allow, deny or approve rules to tool calls. Remote http and sse servers are skipped, since the guard wraps stdio ones.
The single-file CLI needs Node 20 or newer and is MIT-licensed. Free covers verdicts and the guard on your own machine; paid Pro, Team and Enterprise tiers add fleet inventory, Slack approvals, audit logs and SSO. Pricing is per person.
Install
Run in a terminal
claude mcp add killerskills -- npx -y vet-skillsAdd to claude_desktop_config.json
{
"mcpServers": {
"killerskills": {
"command": "npx",
"args": [
"-y",
"vet-skills"
]
}
}
}Add to ~/.cursor/mcp.json
{
"mcpServers": {
"killerskills": {
"command": "npx",
"args": [
"-y",
"vet-skills"
]
}
}
}Add to .vscode/mcp.json
{
"servers": {
"killerskills": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"vet-skills"
]
}
}
}Add to ~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"killerskills": {
"command": "npx",
"args": [
"-y",
"vet-skills"
]
}
}
}Add to your client's MCP settings (most use this shape)
{
"mcpServers": {
"killerskills": {
"command": "npx",
"args": [
"-y",
"vet-skills"
]
}
}
}What it does
- MCP guard
- Wraps stdio servers in 14 agents so every launch is checked.
- Credential placeholders
- Real secrets reach only the vendor's own hosts.
- Host recording
- Notes each host a server contacts and can block new ones by policy.
- Revocation
- Flagged versions refuse to start, and running servers re-check every minute.
- Tool-call rules
- Allow, deny or hold tool calls for approval.
- Pre-install checks
- Verdicts and a honeypot run before a skill or server touches your machine.
Good for
- Check a skill or MCP server before installing it
- Stop a compromised MCP server from stealing environment tokens
- Review which hosts your MCP servers actually contact
- Fail a pull request that adds a blocked skill
Pricing
Freemium
Free for individual developers; Pro, Team, Enterprise and a Verified Publisher plan at $2,400/yr per org
Prices change. Check killerskills.ai before you buy.
KillerSkills FAQ
Is it free?
Yes for individuals, including the guard, verdicts and the GitHub Action; paid tiers add team policy, inventory and audit features.
What does it send to KillerSkills?
Skill name, version, content hash, agent type and the policy outcome, never your code, prompts or file contents.
Does the guard cover remote MCP servers?
No, it wraps stdio servers; remote http and sse servers are skipped.
What does it need?
Node 20 or newer.
Is KillerSkills free?
KillerSkills has a free plan, with paid plans for more usage or features.
How do I add KillerSkills to Claude Code?
Run this command in your terminal: claude mcp add killerskills -- npx -y vet-skills
What are the best alternatives to KillerSkills?
The closest alternatives to KillerSkills are REA, Agent Roast, Midplane, Leash and Tinplate Edge AI Safety, Security & EU Compliance.
Alternatives to KillerSkills

REA
MCP serverCLI and MCP server that lets coding agents reverse engineer apps and binaries
Runs locally- Open source
- 464K GitHub stars
- Developer tools
Agent Roast
MCP serverOffline review that flags agent changes which weaken tests, assertions or CI
Runs locally- Open source
- 4 Hacker News points
- Developer tools

Midplane
MCP serverPostgres gateway for AI agents with table grants, column masking and approval for writes
Runs locally- Open source
- 5 GitHub stars
- Databases

Leash
MCP serverReview what your AI touched, approve risky actions and undo supported changes
- 1 tool
- Remote
- Free
- 1 GitHub star
- Security
Zero-log MCP server with 49 tools for AI guardrails, secret scanning and EU compliance
- 49 tools
- Remote
- Freemium
- Security

SecureCloud
MCP serverSentinelOne and IT security catalog, prices, quotes and checkout links for Israel
- 7 tools
- Remote
- Free
- Security