M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
COMPARE UP TO 4 SERVERS

Compare MCP servers

Compare scores, permissions, risks, and fit in one decision-focused table.

DimensionMCP HandlerEasily spin up an MCP server on Next.js, Nuxt, Svelte, and moreMartinLoop MCP ServerAn independent control layer that gives AI coding agents budgets, brakes, and receipts
FMRS57 / 100 · C79 / 100 · B
Reliability9 / 2012 / 20
Security and permissions11 / 2018 / 20
Maintenance12 / 2016 / 20
Documentation13 / 2018 / 20
Setup experience12 / 2015 / 20
Best for
  • Developers working in JavaScript/TypeScript with Fetch-compatible frameworks
  • Teams wanting one handler that serves both new and legacy MCP clients
  • Stateless, per-request deployments that do not need Redis sessions
  • Projects needing RFC 9728 / RFC 8414 compliant authorization surfaces
  • Engineering teams scaling AI coding agents while controlling spend
  • Platform teams that need auditable, rollback-aware agent execution records
  • Organizations using multiple agents (Claude, Codex, Gemini) under one governed flow
Not for
  • Frameworks built on Node.js IncomingMessage/ServerResponse such as Express without a Web Request adapter
  • Projects still on @modelcontextprotocol/sdk 1.x (use mcp-handler 1.x instead)
  • Deployments requiring the HTTP+SSE (2024-11-05) transport, removed in 2.x
  • Environments running Node.js below 20
  • Developers who just want a simple code-completion helper without run governance
  • Teams unwilling to accept local JSONL run records with HMAC signing
  • Environments without a supported adapter (Claude/Codex/Gemini CLIs, direct provider, or verifier-only)
Required permissions
  • Access to the exposed HTTP route
  • Access to the authorization server that issues bearer tokens
  • Ability to configure protected resource metadata endpoints
  • Network access from the deployment to the authorization server when verifying tokens
  • Read/write access to the configured repo root (writable scope restrictable via --allow-path/--deny-path)
  • Ability to invoke the chosen agent CLI (Claude/Codex/Gemini, etc.) and its underlying model account
  • Local disk writes for run records, receipts, and share outputs (share/run-receipt., etc.)
  • Execution of user-configured verifier commands (e.g. npm test)
Risks and side effects
  • Exposing the MCP handler publicly without auth can let arbitrary callers invoke registered tools
  • Tool return content flows into the model context, so sensitive data could leak
  • 2.x is stateless, so GET/DELETE session operations return 405 and session-dependent clients may misbehave
  • Migrating from 1.x is a breaking change (full Standard Schema for inputSchema, removal of variadic methods, ctx.http?.authInfo), and improper upgrades cause runtime errors
  • Verifier commands are executed; poorly chosen verifier commands carry security risk despite preflight policy checks
  • Cost and token figures may be estimates (labeled with provenance) and should not be treated as settled accounting
  • Receipt integrity must be verified before use as trustworthy evidence; missing rollback evidence is flagged as EVIDENCE_BOUNDARY
  • Policy defaults come from martin.config.yaml and can be overridden by CLI flags; misconfiguration can loosen safety boundaries
Supported clientsClaude Desktop, Cursor, WindsurfClaude Code, Codex, Gemini CLI, Cursor, VS Code, GitHub Copilot, Continue
Tools10