`openclaw mcp` has two jobs:

- run OpenClaw as an MCP server with `openclaw mcp serve`

- manage OpenClaw-owned outbound MCP server definitions with `list`, `show`, `status`, `doctor`, `probe`, `add`, `set`, `configure`, `tools`, `login`, `logout`, `reload`, and `unset`

- manage OpenClaw-managed outbound MCP server definitions with `list`, `show`, `status`, `doctor`, `probe`, `add`, `set`, `configure`, `tools`, `login`, `logout`, `reload`, and `unset`

In other words:

- `serve` is OpenClaw acting as an MCP server

- the other subcommands are OpenClaw acting as an MCP client-side registry for MCP servers its runtimes may consume later

<Note>

`list`, `show`, `set`, and `unset` only read and write OpenClaw-managed `mcp.servers` entries in OpenClaw config. They do not include mcporter servers from `config/mcporter.json`; use `mcporter list` for that registry.

</Note>

Use [`openclaw acp`](/cli/acp) when OpenClaw should host a coding harness session itself and route that runtime through ACP.

## Choose the right MCP path

This is the `openclaw mcp list`, `show`, `status`, `doctor`, `probe`, `add`, `set`,

`configure`, `tools`, `login`, `logout`, `reload`, and `unset` path.

These commands do not expose OpenClaw over MCP. They manage OpenClaw-owned MCP server definitions under `mcp.servers` in OpenClaw config.

These commands do not expose OpenClaw over MCP. They manage OpenClaw-managed MCP server definitions under `mcp.servers` in OpenClaw config. They do not read mcporter servers from `config/mcporter.json`.

Those saved definitions are for runtimes that OpenClaw launches or configures later, such as embedded OpenClaw and other runtime adapters. OpenClaw stores the definitions centrally so those runtimes do not need to keep their own duplicate MCP server lists.

});

});

it("labels listed MCP servers as OpenClaw-managed", async () => {

await withTempHome("openclaw-cli-mcp-home-", async () => {

const workspaceDir = await createWorkspace();

vi.spyOn(process, "cwd").mockReturnValue(workspaceDir);

await runMcpCommand(["mcp", "set", "context7", '{"command":"uvx","args":["context7-mcp"]}']);

mockLog.mockClear();

await runMcpCommand(["mcp", "list"]);

const output = mockLog.mock.calls.map((call) => String(call[0])).join("\n");

expect(output).toContain("OpenClaw-managed MCP servers (");

expect(output).toContain("- context7");

expect(output).toContain("OpenClaw-managed mcp.servers entries");

expect(output).toContain("does not include mcporter servers from config/mcporter.json");

});

});

it("updates per-server MCP tool filters", async () => {

await withTempHome("openclaw-cli-mcp-home-", async (home) => {

const workspaceDir = await createWorkspace();

mockLog.mockClear();

await runMcpCommand(["mcp", "list"]);

expect(lastLogLine()).toContain("No MCP servers configured in ");

const output = mockLog.mock.calls.map((call) => String(call[0])).join("\n");

expect(output).toContain("No OpenClaw-managed MCP servers configured in ");

expect(output).toContain("does not include mcporter servers from config/mcporter.json");

});

});

}

}

const OPENCLAW_MCP_REGISTRY_SCOPE_NOTE =

"Note: this command only shows OpenClaw-managed mcp.servers entries and does not include mcporter servers from config/mcporter.json.";

export function registerMcpCli(program: Command) {

const mcp = program.command("mcp").description("Manage OpenClaw MCP config and channel bridge");

const mcp = program

.command("mcp")

.description("Manage OpenClaw mcp.servers config and channel bridge");

mcp

.command("serve")

mcp

.command("list")

.description("List configured MCP servers")

.description("List OpenClaw-managed MCP servers from mcp.servers")

.option("--json", "Print JSON")

.action(async (opts: { json?: boolean }) => {

const loaded = await listConfiguredMcpServers();

const names = Object.keys(loaded.mcpServers).toSorted();

if (names.length === 0) {

defaultRuntime.log(

`No MCP servers configured in ${loaded.path}. Add one with ${formatCliCommand('openclaw mcp set <name> \'{"command":"uvx","args":["context7-mcp"]}\'')}.`,

`No OpenClaw-managed MCP servers configured in ${loaded.path}. Add one with ${formatCliCommand('openclaw mcp set <name> \'{"command":"uvx","args":["context7-mcp"]}\'')}.`,

);

defaultRuntime.log(OPENCLAW_MCP_REGISTRY_SCOPE_NOTE);

return;

}

defaultRuntime.log(`MCP servers (${loaded.path}):`);

defaultRuntime.log(`OpenClaw-managed MCP servers (${loaded.path}):`);

for (const name of names) {

defaultRuntime.log(`- ${name}`);

}

defaultRuntime.log("");

defaultRuntime.log(OPENCLAW_MCP_REGISTRY_SCOPE_NOTE);

});

mcp

.command("show")

.description("Show one configured MCP server or the full MCP config")

.description("Show one OpenClaw-managed MCP server or the full mcp.servers config")

.argument("[name]", "MCP server name")

.option("--json", "Print JSON")

.action(async (name: string | undefined, opts: { json?: boolean }) => {

return;

}

if (name) {

defaultRuntime.log(`MCP server "${name}" (${loaded.path}):`);

defaultRuntime.log(`OpenClaw-managed MCP server "${name}" (${loaded.path}):`);

} else {

defaultRuntime.log(`MCP servers (${loaded.path}):`);

defaultRuntime.log(`OpenClaw-managed MCP servers (${loaded.path}):`);

}

printJson(value ?? {});

});

mcp

.command("set")

.description("Set one configured MCP server from a JSON object")

.description("Set one OpenClaw-managed MCP server from a JSON object")

.argument("<name>", "MCP server name")

.argument("<value>", 'JSON object, for example {"command":"uvx","args":["context7-mcp"]}')

.action(async (name: string, rawValue: string) => {

mcp

.command("unset")

.description("Remove one configured MCP server")

.description("Remove one OpenClaw-managed MCP server")

.argument("<name>", "MCP server name")

.action(async (name: string) => {

const loaded = await listConfiguredMcpServers();