GitHub - alisaitteke/temporal-mcp: Temporal MCP is a Model Context Protocol (MCP) server for running and managing Temporal workflows from AI assistants. It exposes Temporal operations as MCP tools so agents can start workflows, query state, and send signals without custom integration code. · GitHub
You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Star5 (5)You must be signed in to star a repository
About
Temporal MCP is a Model Context Protocol (MCP) server for running and managing Temporal workflows from AI assistants. It exposes Temporal operations as MCP tools so agents can start workflows, query state, and send signals without custom integration code.
MCP server for the Temporal HTTP API. Manage workflows, namespaces, schedules, and more directly from Cursor, Claude Code, or any MCP-compatible IDE.
Disclaimer: This is an independent, community-built MCP server and is not affiliated with, endorsed by, or officially supported by Temporal Technologies, Inc.
36 tools covering the full Temporal HTTP API surface. By default, only the 11 most essential tools are exposed to keep the LLM context lean — see Tool tiers.
Tools
Cluster
Tool
What it does
get_cluster_info
Server version, cluster ID, and enabled capabilities
Namespaces
Tool
What it does
list_namespaces
List all namespaces in the cluster
describe_namespace
Config, retention period, replication info
Workflows
Tool
What it does
count_workflows
Count executions matching a visibility query
list_workflows
List or search executions using visibility query syntax
describe_workflow
Status, type, task queue, start/close time
start_workflow
Start a new workflow execution
signal_workflow
Send a signal to a running workflow
signal_with_start_workflow
Start + signal in one atomic call (or signal if already running)
query_workflow
Query workflow state via a registered query handler
cancel_workflow
Request graceful cancellation
terminate_workflow
Force-terminate immediately (no cleanup)
pause_workflow
Pause execution (stops scheduling new tasks)
unpause_workflow
Resume a paused workflow
get_workflow_history
Event history with human-readable summaries
Schedules
Tool
What it does
list_schedules
List all schedules in a namespace
describe_schedule
Spec, state, recent actions, next run times
create_schedule
Create a cron or interval-based schedule
delete_schedule
Delete a schedule
Activities
Tool
What it does
list_activities
List activity executions (filter by workflow, type, status)
describe_activity
State, attempt count, heartbeat, failure details
Batch Operations
Tool
What it does
list_batch_operations
List bulk signal/cancel/terminate jobs
describe_batch_operation
Progress, state, success/failure counts
stop_batch_operation
Stop a running batch job
Worker Deployments
Tool
What it does
list_worker_deployments
List all versioned worker deployments
describe_worker_deployment
Current/ramping version, version history
Nexus Endpoints
Tool
What it does
list_nexus_endpoints
List all Nexus endpoints in the cluster
get_nexus_endpoint
Endpoint target namespace and task queue
create_nexus_endpoint
Register a new Nexus endpoint
delete_nexus_endpoint
Remove a Nexus endpoint
Workflow Rules
Tool
What it does
list_workflow_rules
List auto-action rules for matching workflows
describe_workflow_rule
Rule query, action, and creation time
create_workflow_rule
Create a TERMINATE or PAUSE rule with a visibility query
delete_workflow_rule
Delete a rule
Task Queues & Search Attributes
Tool
What it does
describe_task_queue
Active pollers, task ID block
list_search_attributes
Custom and system search attributes
Requirements
Node.js >= 18
A running Temporal server (local or remote)
Add to Cursor
Open or create .cursor/mcp.json in your project root (or ~/.cursor/mcp.json for global):
Key tools return both a human-readable text summary and a machine-readable structuredContent JSON object. Clients that support MCP structured output (e.g. Claude) can use the structured data for further processing without parsing text.
Thanks to everyone who helps improve this project!
About
Temporal MCP is a Model Context Protocol (MCP) server for running and managing Temporal workflows from AI assistants. It exposes Temporal operations as MCP tools so agents can start workflows, query state, and send signals without custom integration code.