Pipeline MCP
Model Context Protocol integration for Pipeline
Pipeline MCP
A data ingestion and transformation service that processes streaming events with SQL and delivers them to object storage in analytics-ready formats.
Installation
npx @dotdo/mcp install pipelineTools
pipeline_create
Create a new data pipeline.
Tool Schema:
{
"name": "pipeline_create",
"description": "Create a new data pipeline.",
"inputSchema": {
"type": "object",
"properties": {
// Tool parameters
}
}
}pipeline_delete
Permanently remove the pipeline.
Tool Schema:
{
"name": "pipeline_delete",
"description": "Permanently remove the pipeline.",
"inputSchema": {
"type": "object",
"properties": {
// Tool parameters
}
}
}pipeline_start
Start processing events through the pipeline.
Tool Schema:
{
"name": "pipeline_start",
"description": "Start processing events through the pipeline.",
"inputSchema": {
"type": "object",
"properties": {
// Tool parameters
}
}
}pipeline_pause
Temporarily stop processing events.
Tool Schema:
{
"name": "pipeline_pause",
"description": "Temporarily stop processing events.",
"inputSchema": {
"type": "object",
"properties": {
// Tool parameters
}
}
}pipeline_resume
Resume a paused pipeline.
Tool Schema:
{
"name": "pipeline_resume",
"description": "Resume a paused pipeline.",
"inputSchema": {
"type": "object",
"properties": {
// Tool parameters
}
}
}pipeline_stop
Stop the pipeline and flush remaining events.
Tool Schema:
{
"name": "pipeline_stop",
"description": "Stop the pipeline and flush remaining events.",
"inputSchema": {
"type": "object",
"properties": {
// Tool parameters
}
}
}pipeline_ingest
Send an event to the pipeline for processing.
Tool Schema:
{
"name": "pipeline_ingest",
"description": "Send an event to the pipeline for processing.",
"inputSchema": {
"type": "object",
"properties": {
// Tool parameters
}
}
}pipeline_transform
Update the SQL transformation query.
Tool Schema:
{
"name": "pipeline_transform",
"description": "Update the SQL transformation query.",
"inputSchema": {
"type": "object",
"properties": {
// Tool parameters
}
}
}pipeline_monitor
Get pipeline metrics (throughput, errors, latency).
Tool Schema:
{
"name": "pipeline_monitor",
"description": "Get pipeline metrics (throughput, errors, latency).",
"inputSchema": {
"type": "object",
"properties": {
// Tool parameters
}
}
}pipeline_logs
Retrieve pipeline execution logs.
Tool Schema:
{
"name": "pipeline_logs",
"description": "Retrieve pipeline execution logs.",
"inputSchema": {
"type": "object",
"properties": {
// Tool parameters
}
}
}Resources
pipeline/created
Triggered when a new pipeline is created.
Resource URI: pipeline://created
pipeline/deleted
Triggered when pipeline is removed.
Resource URI: pipeline://deleted
pipeline/started
Triggered when pipeline begins processing.
Resource URI: pipeline://started
pipeline/paused
Triggered when pipeline is temporarily stopped.
Resource URI: pipeline://paused
pipeline/resumed
Triggered when paused pipeline continues.
Resource URI: pipeline://resumed
pipeline/stopped
Triggered when pipeline is stopped.
Resource URI: pipeline://stopped
pipeline/ingested
Triggered when event is received by pipeline.
Resource URI: pipeline://ingested
pipeline/transformed
Triggered when event is processed by SQL transform.
Resource URI: pipeline://transformed
pipeline/written
Triggered when batch is written to R2.
Resource URI: pipeline://written
pipeline/failed
Triggered when pipeline encounters error.
Resource URI: pipeline://failed
Usage with Claude Desktop
Add to claude_desktop_config.json:
{
"mcpServers": {
"pipeline": {
"command": "npx",
"args": ["-y", "@dotdo/mcp", "serve", "pipeline"]
}
}
}