mcp-open-api
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@mcp-open-apiimport the Petstore OpenAPI spec and propose LLM-ready tools for review"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
mcp-open-api
Turns an OpenAPI spec into a slot of an mcp-broker, validated by a human before it is published, then governed by the broker like any other provider.
OpenAPI spec ──> mcp-open-api (slot "designer") ──> Tier 4 page ──publish──> slot definition (JSON)
│ │ │
└── pushes its page into the └── operator's own token └── run by the broker:
broker's static mount + checks re-run by authorize, audit, limits,
the broker traceparent, secrets keptConverting OpenAPI to MCP is not new. A 300-operation spec does not make a good slot, though: operations have to be chosen, renamed, described for an LLM, narrowed with fixed parameters, and tied to a capability and a resource path. That design work is what this project is about, and what makes a REST API safe to hand to an agent.
How it works
A provider, not a broker feature. mcp-open-api registers on the broker as an ordinary provider, on the slot
designer. Its MCP tools import a spec, propose tools, tune them, validate and try them. An agent can drive them.A human publishes. A transformation becomes a published slot only after a Tier 4 operator validates it, on a web page that mcp-open-api pushes into the broker's static mount. The operator signs in with their own token, reviews the checks and the diff, approves each write tool one by one, and publishes. The broker re-runs every check and audits the publication under the operator's name. The provider itself holds no publishing right.
The page can be a slot. While validating, the page can publish itself as a temporary slot (
preview-<slot>-<id>) exposing the transformed tools, so they can be called for real before publication. Calls are relayed to the broker; API secrets never reach the browser.The output is declarative. A published slot is a slot definition: one readable, diffable JSON file with no secret in it. The broker runs it without generated code and without an extra process. Every call goes through the broker's declared authorization, audit, execution limits and W3C trace propagation, and the target API's credentials stay in the broker's security file.
Related MCP server: MCP Flow
Status
Design stage. The package is scaffolded; no feature is implemented yet. The design covers the slot definition format, the declarative runtime and its governance in the broker, the broker's admin routes and pushed static pages, mcp-open-api itself, the temporary preview slot, and the future MCP path for agents.
Requirements
Node.js 20.11 or later.
An mcp-broker, 1.6.1 or later. Publishing slots and pushed pages need broker features that are not released yet.
Development
npm install
npm run build
npm testLicense
Apache-2.0. See LICENSE.
This server cannot be deployed
Maintenance
Related MCP Connectors
Point Gecko at an OpenAPI spec; get first-call-correct, auth-hidden agent tools.
Scores how well AI agents will use your API, 0-100, from its OpenAPI spec. Free and read-only.
Create hosted MCP servers from any OpenAPI spec. Requires a free Kaiva Bridge account.
Find, vet, and run MCP tools through a secure audited gateway with prompt-injection risk scoring
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceTurns OpenAPI specs into MCP tools with secure defaults, risk inspection, confirmation gates, response limits, audit logging, and secret redaction.-
- FlicenseNot gradedqualityBmaintenanceEnables turning existing REST APIs into governed, agent-callable MCP servers by analyzing OpenAPI schemas, generating tool definitions, and gating releases behind automated evals.-

TCPcore Governance Kernelofficial
AlicenseNot gradedqualityAmaintenanceCompiles YAML adapters or OpenAPI specs into a minimal MCP tool surface that exposes only declared capabilities to agents, with risk-tiered execution, human approval queues, credential brokering, prompt-injection sanitising and NIST AU-3 audit logging. This lets operators put scoped identity, permissions and kill switches in front of any API so agents can call exactly the tools they need and nothing more.3MIT- AlicenseNot gradedqualityAmaintenanceEnables users to turn any OpenAPI 3.x spec into callable MCP tools at runtime, letting agents inspect and invoke API operations directly from JSON or YAML specifications without code generation or build steps.7 npmMIT