AI agents use update_memory to create or update resources in Lara — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Lara environment.
This tool modifies data (translation memory) reversibly. It does not delete, execute code, or move money. Update operations are Write-category risks because they change state but remain undoable via further updates. Severity is medium because mistaken updates to translation memories could corrupt translations used in production, but the blast radius depends on scope and reversibility.
From the tool's definition Tool name is 'update_memory' and description states it 'Updates a translation memory in your Lara Translate account.' The verb 'updates' indicates modification of existing data.
Attacks that exploit this kind of access
Updates a translation memory in your Lara Translate account. It is categorised as a Write tool in the Lara MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Lara MCP server in PolicyLayer and add a rule for update_memory: 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 Lara. Nothing to install.
update_memory 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 update_memory 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 update_memory. 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.
update_memory is provided by the Lara MCP server (@translated/lara-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 →