AI agents invoke set_initial_pose to trigger actions in Ros. 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.
Setting the initial pose on an AMR resets the robot's localization/odometry reference frame, which can cause the robot to misidentify its position and navigate incorrectly. This is an external operation affecting a physical system (via MQTT), not merely reading or writing data in a database. Misuse could send the robot to wrong locations or cause collisions, giving it high severity.
From the tool's definition 'Publish an initial pose and wait for confirmation' — sets the robot's initial pose in its localization system, triggering an external operation on a physical autonomous mobile robot
Attacks that exploit this kind of access
Publish an initial pose and wait for confirmation. It is categorised as a Execute tool in the Ros MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Ros MCP server in PolicyLayer and add a rule for set_initial_pose: 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 Ros. Nothing to install.
set_initial_pose 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 set_initial_pose 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 set_initial_pose. 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.
set_initial_pose is provided by the Ros MCP server (reidlo5135/ros-mcp-server). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →