AI agents invoke container_update to trigger actions in Homelab. 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 actively modifies running infrastructure by pulling new container images and restarting containers (or running docker compose up). It executes operational changes against live systems. While it doesn't permanently destroy data, it can disrupt running services, introduce breaking changes from new images, and affect production-like homelab workloads — making it high severity Execute.
From the tool's definition pull the latest image and restart it... uses docker compose pull + up instead
Attacks that exploit this kind of access
Update a specific container on the devbox: pull the latest image and restart it. If project_dir is provided, uses docker compose pull + up instead. It is categorised as a Execute tool in the Homelab MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Homelab MCP server in PolicyLayer and add a rule for container_update: 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 Homelab. Nothing to install.
container_update 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 container_update 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 container_update. 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.
container_update is provided by the Homelab MCP server (nainounen/homelab-mcp). 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 →