mcpsec

mcpsec is a Python CLI toolchain for security scanning and fuzzing of MCP (Model Context Protocol) servers. It can scan MCP servers via stdio or HTTP, enumerate attack surface, run SAST-like checks (Semgrep rules), and perform runtime fuzzing/generation to find crashes and application-layer vulnerabilities (e.g., SSRF/path traversal/RCE-class indicators), with optional AI-assisted payload generation.

Evaluated Mar 30, 2026 (0d ago)
Repo ↗ Security ai-ml security appsec mcp fuzzing pentesting ssrf injection cli python
⚙ Agent Friendliness
40
/ 100
Can an agent use this?
🔒 Security
42
/ 100
Is it safe for agents?
⚡ Reliability
35
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
20
Documentation
70
Error Messages
0
Auth Simplicity
70
Rate Limits
0

🔒 Security

TLS Enforcement
70
Auth Strength
45
Scope Granularity
0
Dep. Hygiene
55
Secret Handling
45

Tool is designed for offensive testing (fuzzing/exploitation assistance), so operational safety matters. Transport security for HTTP is not clearly specified (assumed HTTPS if using https://). Authentication support appears limited to passing a Bearer token header; scope/granularity and secure secret handling details are not described. Dependency hygiene cannot be verified from the provided manifest alone; listed deps look standard but no CVE review is provided.

⚡ Reliability

Uptime/SLA
0
Version Stability
55
Breaking Changes
50
Error Recovery
35
AF Security Reliability

Best When

You are evaluating your own MCP server/tool ecosystem (including downstream IDE/agent integrations) and can run scans in a controlled/staging environment with safeguards.

Avoid When

You cannot isolate network/file access for the target, cannot handle potential crashes/DoS from fuzzing, or cannot ensure legal/ethical authorization.

Use Cases

  • Assess security of MCP servers you own/have permission to test (runtime scanning, fuzzing, and REPL-style exploitation assistance)
  • Continuously evaluate MCP toolchains for injection/path/SSRF/DoS and related classes of vulnerabilities
  • Generate evidence (PoC/crash cases) and export results to JSON/SARIF for CI workflows
  • Static analysis of MCP codebases using Semgrep rules for common security issues and MCP-specific misconfigurations

Not For

  • Testing MCP servers you do not own or lack explicit permission to test
  • Production exploitation against third parties or unauthorized red-team activity
  • Environments that cannot tolerate active probing (fuzzing can crash servers / create load)

Interface

REST API
No
GraphQL
No
gRPC
No
MCP Server
No
SDK
No
Webhooks
No

Authentication

Methods: HTTP Bearer token via -H/Authorization header for --http scans
OAuth: No Scopes: No

The README indicates HTTP scanning with an Authorization: Bearer TOKEN header, but does not describe OAuth flows, scope granularity, or token lifecycle.

Pricing

Free tier: No
Requires CC: No

Open-source (MIT) CLI; costs are compute/AI-provider usage if using --ai and a configured model provider.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • Active fuzzing/scanning can crash or DoS the target MCP server.
  • HTTP scanning requires correctly formatted MCP endpoint/transport path and valid Authorization header if enabled on the target.
  • AI payload generation requires additional setup (mcpsec setup) and may be sensitive to provider/model configuration.

Alternatives

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for mcpsec.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

Scores are editorial opinions as of 2026-03-30.

6533
Packages Evaluated
19870
Need Evaluation
586
Need Re-evaluation
Community Powered