Skip to content
MCP server

k9 Security Reachable Risk

MCP server that ranks Dependabot CVE alerts by reachability, KEV and EPSS for coding

Answers · sign-in required (OAuth)

The server answered 3 hours ago with a sign-in request for OAuth (HTTP 401), so we could not list its tools.

Install
Pricing
Freemium
Platforms
API, CLI, vscode
Runs
Remotely
Access
Needs sign-in (OAuth)
Version
0.1.0
Listing updated
today

Similar MCP servers

About k9 Security Reachable Risk

k9 Security Reachable Risk is a freemium MCP server for security. It runs as a hosted remote server, so there is nothing to install locally.

Reachable Risk is a remote MCP server that connects to an AI coding agent such as Claude Code or Codex and scores vulnerability alerts from a repository's dependency scanner. Instead of ordering alerts by CVSS severity, it combines whether a CVE appears in VulnCheck KEV, its FIRST EPSS exploit probability, and whether the vulnerable code is reachable from the application's own code paths, with the reachability check performed by the agent in the customer's repository.

The service states that the source code itself stays with the customer: the agent needs read access to the repo and sends only the finding and its reachability result to k9 for scoring. Output is a per-alert verdict (fix, schedule, defer or review) with written reasoning and a report that expresses risk in NIST SP 800-30 terms. Teams describe their deployment context once in a risk-context file, and the same alert can be scored separately for runtime, test and infrastructure contexts.

The product is aimed at engineers who fix vulnerabilities and need to justify deferrals to security teams or auditors. Pricing is metered per alert, so teams with large Dependabot backlogs should check the included volume against their alert count.

Install

Run in a terminal

claude mcp add --transport http mcp https://mcp.k9security.io/mcp

Add to Settings → Connectors → Add custom connector

https://mcp.k9security.io/mcp
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "mcp": {
      "url": "https://mcp.k9security.io/mcp"
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "mcp": {
      "type": "http",
      "url": "https://mcp.k9security.io/mcp"
    }
  }
}

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

{
  "mcpServers": {
    "mcp": {
      "serverUrl": "https://mcp.k9security.io/mcp"
    }
  }
}

Add to Grok Bot desktop app → Connect apps → Add a custom connector

https://mcp.k9security.io/mcp

Enter a Name (for example "k9 Security Reachable Risk"), choose Type: Remote HTTPS, paste this as the Server URL and choose Add plugin. When we checked, this server asked for OAuth sign-in. If its docs give you an OAuth client ID and secret, enter them in the form's OAuth Client ID and Client Secret fields.

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "mcp": {
      "type": "http",
      "url": "https://mcp.k9security.io/mcp"
    }
  }
}

Using several servers? Build one config file with the MCP config generator.

Tools this server provides (2)

Taken from the project documentation; the running server may differ.

score_risk
Scores each alert by combining KEV, EPSS and reachability into a risk score.
gather_risk_context
Walks the codebase with the agent and drafts the risk context file.

What it does

Reachability-based alert scoring
The coding agent checks whether each vulnerable dependency is reachable in the application's code, and that result feeds the verdict.
KEV and EPSS signals
Each alert is checked against VulnCheck KEV for known exploitation and FIRST EPSS for 30-day exploit probability.
Per-alert verdicts with reasoning
Every fix, schedule, defer or review decision comes with a written rationale and a trigger that would change it.
Multi-context scoring
One alert can be scored separately for runtime, test and deployment infrastructure contexts based on a risk-context file.
NIST SP 800-30 reporting
Results report inherent, residual and retired risk in the qualitative and semi-quantitative forms used by security teams.
Source code stays local
The agent reads the repository while k9 receives only the finding and reachability result for scoring.

Good for

  • Triaging a large Dependabot backlog to find the few alerts that need a fix this week
  • Writing defensible defer decisions for auditors or customer security reviews
  • Checking whether a Critical advisory touches code the application actually loads
  • Deciding which dependency upgrades to bundle into routine maintenance

Pricing

Freemium, from $10 a month

7-day free trial, then $10/month for 1,000 alerts; Teams plan $150/month listed as coming soon

Prices change. Check k9security.io before you buy.

k9 Security Reachable Risk FAQ

How do I connect it?

The listing gives a three-line configuration that adds the remote MCP server at https://mcp.k9security.io/mcp to Claude Code, Codex or a compatible agent.

Does k9 receive my source code?

The website says k9 never sees source code. The agent needs read access to the repository and sends only the finding and its reachability result for scoring.

What does it cost?

After a 7-day free trial the Pro plan is $10 per month, including 1,000 alerts, with $0.03 per additional alert. A Teams plan at $150 per month is described as coming soon.

Which vulnerability sources does it accept?

The site says findings can come from Dependabot, Wiz, Tenable and similar scanners.

Is k9 Security Reachable Risk free?

k9 Security Reachable Risk has a free plan, with paid plans for more usage or features. Paid plans start at $10 a month.

How do I add k9 Security Reachable Risk to Claude Code?

Run this command in your terminal: claude mcp add --transport http mcp https://mcp.k9security.io/mcp

What are the best alternatives to k9 Security Reachable Risk?

The closest alternatives to k9 Security Reachable Risk are vulnerable.now, Threatist, Patch Priority Watch, Vdb and Release Alert.

Where this comes from

  • Description: written from the project's website and its MCP Registry entry.
  • Status: checked automatically by AIPoweredBot on 19 November 2026.
  • Install steps: from the official MCP Registry entry.

How AIPoweredBot checks servers