OpenAPI MCP
Click on "Install 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., "@OpenAPI MCPuse the petstore API to list all available pets"
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.
OpenAPI MCP
A generic MCP server that builds its tools at runtime from any OpenAPI specification.
Documentation | Config wizard | PyPI | Docker
Features
openapi-mcp builds its MCP tools at runtime from any OpenAPI specification,
with no per-API code. Point one container image at a spec (URL or mounted file)
plus upstream credentials, and it exposes the operations of that API as MCP
tools via FastMCP's OpenAPIProvider.
Intended for simple APIs. Large or complex APIs, or ones using
oauth2/openIdConnect/mutualTLS upstream auth, are better served by a
purpose-built sibling. See
docs/superpowers/specs/2026-07-04-openapi-generic-wrapper-design.md
for the design and .env.example for the full OAPI_* contract.
docker run --rm \
-e OAPI_SPEC_URL=https://api.example.com/openapi.json \
-e OAPI_SECURITY_APIKEYAUTH=your-key \
ghcr.io/pvliesdonk/openapi-mcpRelated MCP server: OpenAPI MCP Server
What you can do with it
The tool surface is whatever the mounted OpenAPI spec defines: each operation
in the spec becomes one MCP tool, named by its operationId. The concrete
prompts depend on the API you wrap. With a spec mounted, you can ask Claude to:
Call an operation directly: "List the open orders" runs the spec's
list_ordersoperation and returns the response.Chain operations: "Find the customer named Acme, then show their most recent invoice" composes a search operation with a lookup operation.
Confirm the deployment: "Which server version is running?" calls the built-in
get_server_infotool.
Because tools are generated from the spec, the exact tool names and working prompts match your API. Point the server at a different spec and the tool surface changes with no code edits.
Installation
From PyPI
pip install pvliesdonk-openapi-mcpIf you add optional extras via the PROJECT-EXTRAS-START / PROJECT-EXTRAS-END sentinels in pyproject.toml, document them below:
From source
git clone https://github.com/pvliesdonk/openapi-mcp.git
cd openapi-mcp
uv sync --all-extras --all-groupsDocker
docker pull ghcr.io/pvliesdonk/openapi-mcp:latestA compose.yml ships at the repo root as a starting point. Copy .env.example to .env, edit, and docker compose up -d.
To attach a remote Python debugger (development only; the protocol is unauthenticated), see Remote debugging.
Linux packages (.deb / .rpm)
Download .deb or .rpm packages from the GitHub Releases page. Both install a hardened systemd unit; env configuration is sourced from /etc/openapi-mcp/env (copy from the shipped /etc/openapi-mcp/env.example).
Claude Desktop (.mcpb bundle)
Download the .mcpb bundle from the GitHub Releases page and double-click to install, or run:
mcpb install openapi-mcp-<version>.mcpbClaude Desktop prompts for required env vars via a GUI wizard, with no manual JSON editing needed.
For manual Claude Desktop configuration and setup options, see Claude Desktop deployment.
Quick start
openapi-mcp serve # stdio transport
openapi-mcp serve --transport http --port 8000 # streamable HTTPFor library usage (embedding the domain logic without the MCP transport), import from the openapi_mcp package directly. See the project's domain modules under src/openapi_mcp/ for entry points.
Server info
The server registers a built-in get_server_info tool (via fastmcp_pvl_core.register_server_info_tool) so operators can confirm the deployed version with a single MCP call. The default response carries server_name, server_version, and core_version. Servers that talk to a remote upstream wire upstream version reporting inside the DOMAIN-UPSTREAM-START / DOMAIN-UPSTREAM-END sentinel in src/openapi_mcp/server.py; see CLAUDE.md for the wiring pattern.
Configuration
Core environment variables shared across all fastmcp-pvl-core-based services:
Variable | Default | Description |
|
| Log level for FastMCP internals and app loggers ( |
|
| Set to |
|
| Persistent-state backend URL for pvl-core subsystems: |
Domain-specific variables go below under Domain configuration.
Authentication
Callers authenticate via a bearer token or OIDC (mutually exclusive). See the Authentication guide for setup, mapped multi-subject tokens, OIDC, and troubleshooting.
Post-scaffold checklist
After copier copy and gh repo create --push:
Fill in the DOMAIN blocks (every section marked with a
DOMAINsentinel comment) in this README and inCLAUDE.md.Configure GitHub secrets (see below).
Install dev + docs tooling:
uv sync --all-extras --all-groups.Install pre-commit hooks:
uv run pre-commit install.Run the gate locally:
uv run pytest -x -q && uv run ruff check --fix . && uv run ruff format . && uv run mypy src/ tests/.Push the first commit. CI should be green.
GitHub secrets
CI workflows reference three repository secrets. Configure them via Settings → Secrets and variables → Actions or with gh secret set:
Secret | Used by | How to generate |
|
| Fine-grained PAT at https://github.com/settings/personal-access-tokens/new with |
|
| https://codecov.io: sign in with GitHub, add the repo, copy the upload token from the repo settings page. |
|
| Run |
gh secret set RELEASE_TOKEN
gh secret set CODECOV_TOKEN
gh secret set CLAUDE_CODE_OAUTH_TOKENGITHUB_TOKEN is auto-provided; no action needed.
Local development
The PR gate (matches CI):
uv run pytest -x -q # tests
uv run ruff check --fix . && uv run ruff format . # lint + format
uv run mypy src/ tests/ # type-checkPre-commit runs a subset of the gate on each commit; see .pre-commit-config.yaml for details, or CLAUDE.md for the full Hard PR Acceptance Gates.
Troubleshooting
Moving a scaffolded project
uv sync creates .venv/bin/* scripts with absolute shebangs pointing at the venv Python. If you move the repo after scaffolding (mv /old/path /new/path), uv run pytest fails with ModuleNotFoundError: No module named 'fastmcp' because the stale shebang resolves to a different interpreter than the venv's site-packages.
Fix:
rm -rf .venv
uv sync --all-extras --all-groupsuv run python -m pytest also works as a one-shot workaround (bypasses the stale entry-script shim).
uv.lock refresh after copier update
When copier update introduces new dependencies (such as a new extra added to pyproject.toml.jinja), CI runs uv sync --frozen which fails against a stale lockfile. Run uv lock locally and commit the refreshed uv.lock alongside accepting the copier-update PR.
Links
Domain configuration
Domain environment variables use the OAPI_ prefix. openapi-mcp derives its
tools from an OpenAPI spec at boot, so configure the spec source, upstream base
URL, timeout, and per-scheme upstream credentials. See
.env.example for a copy-paste template and the
configuration guide
for the full contract.
Variable | Default | Required | Description |
| (none) | Exactly one of URL/PATH | URL of the OpenAPI spec, fetched at boot. |
| (none) | Exactly one of URL/PATH | Local or mounted spec file (JSON or YAML). |
| spec | No | Override the upstream base URL. |
|
| No | Upstream request timeout in seconds. |
| (none) | If the scheme is referenced | Credential for a referenced security scheme, named by its uppercased key. |
Setting both OAPI_SPEC_URL and OAPI_SPEC_PATH, or neither, is a fail-loud
boot error. Upstream credentials cover apiKey (header or query), http
bearer, and http basic (a user:pass value); oauth2, openIdConnect,
and mutualTLS are out of scope and belong in a purpose-built sibling.
Domain-config fields are composed inside src/openapi_mcp/config.py between the CONFIG-FIELDS-START / CONFIG-FIELDS-END sentinels; env reads go through fastmcp_pvl_core.env(_ENV_PREFIX, "SUFFIX", default) so naming stays consistent.
Key design decisions
Tools are derived from the spec at runtime. Each spec operation becomes one MCP tool through FastMCP's
OpenAPIProvider, so pointing the server at a different spec changes the tool surface with no code edits.Two auth layers stay strictly separate. Inbound auth (who may call this server, inherited from
fastmcp-pvl-core) and upstream auth (OAPI_SECURITY_*, how this server authenticates to the wrapped API) never mix.Upstream credentials ride an
httpx.Authflow. They are injected when the request is dispatched rather than as client-level headers or query parameters, so they survive the request constructionOpenAPIProviderperforms and reach the wire for both header and query schemes.Boot fails loud. A missing or malformed spec, an ambiguous spec source, a missing credential, or a non-positive timeout raises at startup instead of surfacing as a per-request error later.
Scope is deliberately narrow. Only
apiKey,http bearer, andhttp basicupstream schemes are supported, and a credential is required for every referenced scheme (no "scheme A or B" alternatives). Larger or complex APIs,oauth2/openIdConnect/mutualTLSauth, and mutually exclusive auth alternatives belong in a purpose-built sibling.
See
docs/superpowers/specs/2026-07-04-openapi-generic-wrapper-design.md
for the full design rationale.
This server cannot be installed
Maintenance
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/pvliesdonk/openapi-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server