Workflows MCP
未认领MCP Server can execute workflows and complex pipelines, similar to Argo or Github Workflow
01 / 选择客户端
03 / 添加配置
claude mcp add --transport stdio 'workflows-mcp' -- 'uvx' 'workflows-mcp'04 / 在客户端中确认
打开客户端的 MCP 设置，确认服务器已连接并列出工具。复制配置不代表连接成功。
更多Automation服务器
浏览完整目录安全概况
已认领和已认证的服务器每周接受一次静态扫描，显示代码能触及的范围（外部服务、环境变量、shell 命令、代理配置目录）以及带有已知公告的依赖。认领此列表即可获得。 安全概况的工作原理
19 个工具中显示 19 个
文档中列出的工具 (19)
内容来自项目文档。服务器握手不会验证每个工具的说明或行为。
cancel_job
Stop queued/running jobs
execute_inline_workflow
Test one-off YAML without registering
execute_workflow
Run a registered workflow
get_job_status
Poll a specific async job
get_queue_stats
Monitor queue health/capacity
get_workflow_info
Exploratory to confirm inputs/outputs
get_workflow_schema
Debugging only. Retrieve full JSON schema for authoring
graph_upsert
with graph.kind="place": all four fields must resolve.
ingest
all four fields must resolve (including compartment).
list_jobs
Find jobs by status (especially paused)
list_workflows
List registered workflows
memory
Unified memory query/ingest/maintenance/graph operations
project_onboard
Current memory onboarding flow with checkpoints
project_sync
Current memory checkpoint resume/continuation
query
all four fields must resolve.
reload_workflows
After editing workflow YAML files on disk
resume_workflow
Continue paused Prompt workflows
scope_token
resolves from execution context memoryscopetokens; contextid resolves from memorycontextscopes.
validate_workflow_yaml
Validate YAML before execution