mcp-gateway

Provides a deployable MCP Gateway (reverse-proxy via Nginx) and a MCP Registry/UI to register and govern MCP servers, discover tools (including schemas via ListTools), and route MCP client traffic to enabled backend MCP servers using per-service URL paths. Includes health checks and real-time UI updates via WebSockets.

Evaluated Mar 30, 2026 (21d ago)
Repo ↗ Infrastructure ai-agents mcp mcp-gateway mcp-registry nginx fastapi sse websockets tool-discovery
⚙ Agent Friendliness
49
/ 100
Can an agent use this?
🔒 Security
47
/ 100
Is it safe for agents?
⚡ Reliability
35
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
70
Documentation
55
Error Messages
0
Auth Simplicity
80
Rate Limits
20

🔒 Security

TLS Enforcement
70
Auth Strength
45
Scope Granularity
20
Dep. Hygiene
40
Secret Handling
60

TLS is supported via HTTPS with cert/key mounting, but README also states self-signed cert default and HTTP is available; this reduces assurance if misconfigured. Authentication for registry UI/API uses username/password (ADMIN_USER/ADMIN_PASSWORD); no evidence of OAuth/OIDC, fine-grained authorization, or scope-based access. Project includes a large dependency list (including ML libraries), but no CVE/lockfile/security posture details are provided in the excerpt. Secrets are passed via environment variables (e.g., ADMIN_PASSWORD, POLYGON_API_KEY, SECRET_KEY) and not shown being logged, but the excerpt does not confirm no logging of secrets.

⚡ Reliability

Uptime/SLA
20
Version Stability
30
Breaking Changes
40
Error Recovery
50
AF Security Reliability

Best When

You want to standardize and govern MCP server access behind a single endpoint with path-based routing and you can operate the gateway/registry stack.

Avoid When

You need turnkey SaaS or vendor-managed scaling and SLAs, or you cannot risk that gateway self-signed cert defaults are used without proper TLS configuration.

Use Cases

  • Centralized discovery and management of approved MCP servers for agent tool use
  • Enterprise/governed access to curated MCP servers across teams or environments
  • Dynamic tool discovery in agents (query tools by natural language) and invocation via the gateway
  • Health monitoring of MCP services (SSE endpoint checks) and enabling/disabling services at runtime

Not For

  • Production use where you cannot operate/manage your own infrastructure (Docker, Nginx, optional EC2/EKS)
  • Environments requiring OAuth/OIDC federation rather than a built-in username/password login
  • Use cases needing fine-grained per-tool authorization beyond what the gateway/registry provides

Interface

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

Authentication

Methods: Basic/battery-included admin login for registry UI and API (ADMIN_USER/ADMIN_PASSWORD) Optional upstream authorization for backend MCP servers (example notes mention 401 if Polygon API key missing)
OAuth: No Scopes: No

Auth is described for securing the registry UI and API using ADMIN_USER/ADMIN_PASSWORD. No evidence in provided content of OAuth/OIDC, scoped tokens, or per-route authorization beyond enable/disable and service management controls.

Pricing

Free tier: No
Requires CC: No

Self-hosted open-source project; costs are infrastructure-related (EC2/EKS, load balancers, TLS certs, etc.).

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • Because it relies on per-service path routing, agents must use the correct gateway base URL and service path (e.g., /weather/, /currenttime/) to reach the intended MCP server.
  • TLS defaults mention self-signed certs and HTTP availability; agents may fail verification unless configured to trust the cert or use proper HTTPS setup.
  • Health checks rely on /sse HEAD probes; a service may be 'enabled' but still fail tool calls if underlying dependencies (e.g., Polygon API key) are misconfigured.
  • Nginx configuration is dynamically regenerated based on enabled/disabled state; changes may require time for reload/regeneration before agents see updates.

Alternatives

Full Evaluation Report

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

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.

8642
Packages Evaluated
17761
Need Evaluation
586
Need Re-evaluation
Community Powered