AI agents call memory_compact to permanently remove resources in Locus — typically in cleanup and lifecycle workflows. It does its job in a single call, and there is no undo.
'Compact' operations typically consolidate or remove data irreversibly (e.g., compacting a database removes deleted records permanently). Given the context of siblings like 'memory_purge' and 'memory_forget', this tool likely performs some form of irreversible data reduction or cleanup. However, the empty description significantly lowers confidence — it could also be a non-destructive optimization.
From the tool's definition Tool name 'memory_compact' on a server with siblings including 'memory_purge' and 'memory_forget', suggesting destructive memory operations; description is empty.
Documented attack patterns abuse exactly the kind of access memory_compact gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Locus, and nothing reaches the server without passing your rules. This is the rule we recommend for memory_compact:
{
"version": "1",
"default": "deny",
"hide": [
"memory_compact"
]
} memory_compact disappears from the agent's tool list entirely, and any attempt to call it is denied. The rest of the server keeps working.
Free to start. No card required.
memory_compact. It is categorised as a Destructive tool in the Locus MCP Server, which means it can permanently delete or destroy data. Block by default and require explicit approval.
Register the Locus MCP server in PolicyLayer and add a rule for memory_compact: 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 Locus. Nothing to install.
memory_compact is a Destructive tool with critical risk. Critical-risk tools should be blocked by default and only enabled with explicit human approval.
Yes. Add a rate_limit block to the memory_compact 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 memory_compact. 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.
memory_compact is provided by the Locus MCP server (magnifico4625/locus). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Locus, 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.
17 Locus tools catalogued and risk-classified — across an index of 43,000+ MCP servers.