workflow_run
Run a workflow from a template or file Use when native TodoWrite + sequential Bash is wrong because the work has a real dependency graph that needs persistence, retry policy, pause/resume, and step-output binding across LLM-driven steps. For a single linear todo list, native TodoWrite is fine.
This record as markdown: /tools/io-github-ruvnet-claude-flow/workflow-run.md
What workflow_run does on Claude Flow
AI agents invoke workflow_run to trigger actions in Claude Flow. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call: builds kicked off, notifications sent, workflows started.
Why workflow_run is rated High
This tool executes workflows which can contain arbitrary LLM-driven steps with dependencies, retries, and state management. While the description does not explicitly state code execution, the capability to run multi-step workflows with LLM integration and dependency graphs allows triggering of external operations with side effects that depend on user-supplied workflow definitions.
From the tool's definition Tool description states "Run a workflow" with "retry policy, pause/resume, and step-output binding across LLM-driven steps" and mentions it handles complex dependency graphs with persistence.
Attacks that exploit this kind of access
The rule that runs workflow_run safely
PolicyLayer is an MCP gateway: it sits between your AI agents and Claude Flow, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For workflow_run, this is the rule to start with:
workflow_run stays usable, but rate-capped: a runaway agent can't fire it dozens of times a minute. Everything else on the server is denied unless you say otherwise.
The button opens the PolicyLayer dashboard: create your workspace, connect Claude Flow, apply this rule, and every workflow_run call is checked against it from then on.
Questions about workflow_run
Run a workflow from a template or file Use when native TodoWrite + sequential Bash is wrong because the work has a real dependency graph that needs persistence, retry policy, pause/resume, and step-output binding across LLM-driven steps. For a single linear todo list, native TodoWrite is fine. It is categorised as a Execute tool in the Claude Flow MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Claude Flow MCP server in PolicyLayer and add a rule for workflow_run: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Claude Flow. Nothing to install.
workflow_run is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the workflow_run rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for workflow_run. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
workflow_run is provided by the Claude Flow MCP server (claude-flow). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
More on Claude Flow, and thousands of servers like it.
Across the catalogue