ymlr-ai-mcp
v0.0.2
Published
Model context protocol for AI in ymlr
Downloads
5
Maintainers
Readme
ymlr-ai-mcp
Model context protocol for AI which is used in ymlr
Shares
Tag details
| Tags | Description | |---|---| | ymlr-ai-mcp'client | Create a MCP Client to call a tool | | ymlr-ai-mcp'server | Create a MCP Server | | ymlr-ai-mcp'tool | Register a MCP Tool handler |
ymlr-ai-mcp'client
Create a MCP Client to call a tool
Example:
Use MCP Client to call a tool "add(a: number, b: number)"
- ymlr-ai-mcp'callTool:
server:
name: My App
version: 1.0.1
sseURL: http://localhost:3000/sse
name: add
params:
a: 1
b: 2
vars: result
- echo: ${ $v.result }ymlr-ai-mcp'server
Create a MCP Server
Example:
Publish a message to redis
- ymlr-ai-mcp'server:
name: My App
version: 1.0.1
opts:
transportType: sse
sse: { endpoint: "/sse", port: 3000 }ymlr-ai-mcp'tool
Register a MCP Tool handler
Example:
Publish a message to redis
- ymlr-ai-mcp'server:
name: My App
version: 1.0.1
opts:
transportType: sse
sse: { endpoint: "/sse", port: 3000 }
runs:
- ymlr-ai-mcp'tool:
name: add
description: Add two numbers
params:
a:
type: number
b:
type: number
runs:
- name: MCP request is ${ JSON.stringify($ps.mcpRequest) }
js: $ps.mcpRequest.response = ($ps.mcpRequest.params.a + $ps.mcpRequest.params.b) + ''Declare out of ymlr-ai-mcp'server
- id: mcpServer
detach: true
ymlr-ai-mcp'server:
name: My App
version: 1.0.1
opts:
transportType: stdio
runs: []
- ymlr-ai-mcp'tool:
mcpServer: ${ $v.mcpServer }
name: add
description: Add two numbers
params:
a:
type: number
b:
type: number
runs:
- name: MCP request is ${ JSON.stringify($ps.mcpRequest) }
js: $ps.mcpRequest.response = ($ps.mcpRequest.params.a + $ps.mcpRequest.params.b) + ''