Request cancellation for one running workflow agent.
AI agents invoke workflow_stop_agent to trigger actions in Codex Workflows MCP Server. 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.
This tool does not merely read data (Read), nor does it permanently destroy data (Destructive). It does not involve financial transactions (Financial). While it modifies state by stopping an agent, the action is reversible through workflow_restart_agent.
From the tool's definition The tool performs an action that triggers an external operation: 'Request cancellation for one running workflow agent.' This directly affects running processes/agents, which is an executable operation with side effects dependent on which agent is targeted.
Documented attack patterns abuse exactly the kind of access workflow_stop_agent gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Codex Workflows MCP Server, and nothing reaches the server without passing your rules. This is the rule we recommend for workflow_stop_agent:
{
"version": "1",
"default": "deny",
"tools": {
"workflow_stop_agent": {
"limits": [
{
"counter": "workflow_stop_agent_rate",
"window": "minute",
"max": 10,
"scope": "grant"
}
]
}
}
}workflow_stop_agent 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.
Free to start. No card required.
Request cancellation for one running workflow agent. It is categorised as a Execute tool in the Codex Workflows MCP Server MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Codex Workflows MCP Server MCP server in PolicyLayer and add a rule for workflow_stop_agent: 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 Codex Workflows MCP Server. Nothing to install.
workflow_stop_agent 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_stop_agent 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_stop_agent. 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_stop_agent is provided by the Codex Workflows MCP Server MCP server (robzilla1738/codex-workflows). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 10 Codex Workflows MCP Server tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
10 Codex Workflows MCP Server tools catalogued and risk-classified — across an index of 42,500+ MCP servers.