Universal MCP Bridge & Proxy
Related Servers
Alternatives to Universal MCP Bridge & Proxy
No user-submitted related servers found.
Related Servers
- AlicenseNot gradedqualityCmaintenanceEnables users to turn any API, CLI, database, website, SDK, or codebase into a production-ready MCP server with secure, dynamically discoverable capabilities.MIT
- AlicenseNot gradedqualityCmaintenanceEnables developers to expose any dev tool as a safe, audited MCP server with permission tiers, audit logging, and automatic argument validation.MIT
- AlicenseAqualityCmaintenanceExposes internal databases, HTTP APIs, and files to LLM agents over MCP, with auto-generated schemas, access controls, rate limiting, redaction, and audit logging.3MIT
- AlicenseNot gradedqualityCmaintenanceOne command to turn any codebase into an MCP server. Not just REST APIs. Not just OpenAPI specs.56Apache 2.0
- AlicenseNot gradedqualityDmaintenanceConvert any OpenAPI spec into a secure MCP server with scoped auth, per-tool allow/deny policies, rate limiting, and a redacted audit trail.10 npmMIT
- AlicenseNot gradedqualityCmaintenanceTurns existing backend APIs (OpenAPI) and SQL databases (PostgreSQL, MariaDB, ClickHouse) into hosted MCP servers for AI clients, with built-in dashboards, forms, and credential encryption.5,309 npmApache 2.0
TDQS
Scored across 7 tools
Each tool targets a clearly different operation: pet lookup by status vs. by ID, text analysis, system health, DB schema inspection, DB query, and curl execution. The only mild overlap is findPetsByStatus vs getPetById, but list-vs-get semantics distinguish them. The bigger problem is thematic incoherence, not tool-to-tool confusion.
The set mixes camelCase (findPetsByStatus, getPetById) with snake_case (run_text_analyzer, db_sample_schema, exec_curl) and uses inconsistent verb styles (find/get/run/exec). Names are individually readable but the conventions change from tool to tool.
Seven tools is a reasonable count in isolation and none feels redundant. However, they are an unrelated grab-bag rather than a scoped surface, so the count is fine but the composition is arbitrary.
There is no coherent domain to cover: pet tools lack create/update/delete, the DB exposes only read-only sample/query with no schema mutation, and text/sys/curl are one-off utilities. As a 'universal bridge' the surface is a random sampling with major lifecycle gaps.