AI agents use complete_task to create or update resources in MCP Orchestrator Server — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your MCP Orchestrator Server environment.
This tool updates task status to a completed state, which is a reversible modification of data. It does not delete data (would be Destructive), execute external code (would be Execute), move money (would be Financial), or merely read data (would be Read). Severity is low because task status updates have limited blast radius and can typically be reversed or corrected by updating the task again.
From the tool's definition Tool name 'complete_task' and description 'Mark a task as completed' indicate a state modification operation on an existing task record.
Documented attack patterns abuse exactly the kind of access complete_task gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and MCP Orchestrator Server, and nothing reaches the server without passing your rules. This is the rule we recommend for complete_task:
{
"version": "1",
"default": "deny",
"tools": {
"complete_task": {
"limits": [
{
"counter": "complete_task_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}complete_task stays usable, but capped — an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Mark a task as completed. It is categorised as a Write tool in the MCP Orchestrator Server MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the MCP Orchestrator Server MCP server in PolicyLayer and add a rule for complete_task: 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 MCP Orchestrator Server. Nothing to install.
complete_task is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the complete_task 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 complete_task. 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.
complete_task is provided by the MCP Orchestrator Server MCP server (mokafari/orchestrator-server). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from MCP Orchestrator Server, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
7 MCP Orchestrator Server tools catalogued and risk-classified — across an index of 43,000+ MCP servers.