Traceforce – AI Security Platform for Devices and MCP Servers

TL;DR

TL;DR: Traceforce is an AI security platform that monitors AI apps, MCP servers, and CLI agents across your device fleet, with an open-source MCP X-Ray scanner that generates SARIF reports for CI/CD integration.

Source and Accuracy Notes

⚠️ This section is MANDATORY. All links must be verified from actual source, not guessed.

• Project page: traceforce.ai ← MUST visit and verify
• Source repository: github.com/traceforce/mcp-xray ← MUST read README
• License: Apache-2.0 (verified via GitHub API license.spdx_id)
• Atlas registry: atlas.traceforce.co — 600+ MCPs catalogued
• HN launch thread: news.ycombinator.com/item?id=48937020
• Source last checked: 2026-08-01 (commit verified against GitHub)

What Is Traceforce?

Traceforce is an on-device AI security platform built for enterprises and teams deploying AI tools at scale. It monitors and controls AI applications, MCP servers, and CLI agents from a centralized dashboard with a claimed 15-minute setup time.

The platform has two components:

1. Traceforce Enterprise — A cloud-hosted or self-hosted control plane that gives security teams visibility across the entire device fleet. It provides automated remediation, real-time threat blocking, and compliance tracking for AI tool usage.

2. MCP X-Ray — A fully open-source security scanner for Model Context Protocol servers. It generates SARIF-formatted reports suitable for CI/CD pipeline integration, scanning both MCP configurations and live server behavior.

The Atlas registry at atlas.traceforce.co maintains security profiles for over 600 MCP servers — giving teams a head start on risk assessment without scanning from scratch.

MCP X-Ray: Open-Source MCP Security Scanner

MCP X-Ray is the open-source portion of Traceforce’s offering. It targets developers and security engineers working with MCP servers in Cursor, Claude Desktop, or Windsurf.

Installation

MCP X-Ray requires Go 1.25.4 or later.

git clone https://github.com/traceforce/mcp-xray
cd mcp-xray
make install-dependencies
make all

The binary is created as ./mcpxray in the current directory.

Config Scan

Scans MCP configuration files for security issues before running any pentest. It uses the Token Analyzer by default, or an LLM-based analyzer for deeper analysis.

# Scan a specific MCP config file
./mcpxray config-scan /path/to/mcp/config.json

# Scan all known config paths automatically (Cursor, Claude, Windsurf)
./mcpxray config-scan --scan-known-configs

# Use LLM analyzer for extensive analysis
./mcpxray config-scan /path/to/mcp/config.json \
  --analyzer-type llm \
  --llm-model claude-3-5-sonnet-20241022

# Output SARIF report
./mcpxray config-scan /path/to/mcp/config.json \
  --output mcp-security-report.sarif.json

Config scan detection capabilities:

• Connection Security — Validates TLS certificates, detects unsafe localhost/loopback exposure, validates OAuth 2.0 configuration (PRM/ASMD)
• Secrets Detection — Scans for exposed credentials, API keys, and other sensitive information
• Tool Analysis — Analyzes tool descriptions for risks including arbitrary execution, injection vulnerabilities, authorization bypass, and information disclosure

Pentest

Executes actual tool calls against MCP servers using auto-generated test plans. Requires an LLM model.

# Run pentest with auto-generated test plan
./mcpxray pentest /path/to/mcp/config.json \
  --llm-model claude-sonnet-4-5

# Use a custom test plan YAML file
./mcpxray pentest /path/to/mcp/config.json \
  --test-plan /path/to/test-plan.yaml \
  --llm-model claude-sonnet-4-5

Pentest detection capabilities: Code execution, SSRF, path traversal, authorization bypass, input injection, information disclosure, and DoS vulnerabilities.

Scan results can be uploaded to Traceforce Atlas for centralized security management and tracking.

Traceforce Enterprise Platform

The commercial platform adds fleet-wide monitoring on top of the MCP scanner. Security teams get:

• Real-time AI activity monitoring across all enrolled devices
• Automated remediation when risky AI tool behavior is detected
• Compliance posture tracking for regulated industries
• MCP registry integration — Atlas profiles for 600+ MCPs pre-loaded

The platform is designed for enterprise security teams that need to audit AI tool usage across developer machines without blocking developers from using AI assistants.

Practical Evaluation Checklist

• [ ] Clone and build mcpxray from source (Go 1.25.4+ required)
• [ ] Run config-scan --scan-known-configs on your own workstation
• [ ] Inspect the SARIF output and verify it integrates with your CI/CD tooling
• [ ] If on an enterprise plan, enroll a test device and confirm fleet visibility
• [ ] Compare Atlas registry entries against MCPs you actively use

Security Notes

• MCP X-Ray makes actual tool calls against MCP servers during pentest — do not run against production servers without explicit permission
• Config scan is read-only and safe to run against any config file
• SARIF output can be fed into GitHub Advanced Security, Semgrep, or any SARIF-compatible analysis pipeline
• The Atlas registry is community-curated — treat individual MCP security profiles as one input among several, not as authoritative verdicts

FAQ

Q: Is MCP X-Ray free to use? A: Yes. MCP X-Ray is fully open-source under the Apache-2.0 license. The Traceforce Enterprise platform is a separate commercial product.

Q: Which MCP servers does it support? A: MCP X-Ray works with any MCP server that follows the standard MCP protocol. The Atlas registry has pre-built security profiles for over 600 MCPs.

Q: What is SARIF and why does it matter? A: SARIF (Static Analysis Results Interchange Format) is an OASIS standard format for sharing static analysis results. MCP X-Ray’s SARIF output integrates with GitHub Advanced Security, Azure DevOps, and most security scanning pipelines.

Q: Does Traceforce support self-hosted deployment? A: The MCP X-Ray scanner runs entirely locally. The Traceforce Enterprise platform supports cloud-hosted and self-hosted deployment options for enterprise customers.

Conclusion

Traceforce sits at the intersection of AI tool adoption and enterprise security — a gap that widens as AI assistants and MCP servers proliferate across developer workstations. MCP X-Ray gives individual developers a free, open-source way to audit MCP configurations and catch common security mistakes before they reach production. The commercial platform scales that visibility to security teams managing hundreds of devices.

The 15-minute setup claim applies to the enterprise platform enrollment. MCP X-Ray from source takes a few minutes to build if you already have Go installed, longer if you need to install the toolchain first.

If your team uses Cursor, Claude Desktop, or Windsurf with MCP servers and you do not have a formal review process for those integrations, running mcpxray config-scan --scan-known-configs is a quick win.