AI agents use invalidate_fact to create or update resources in Memoraeu — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Memoraeu environment.
The tool modifies the state of an existing fact by marking it as expired. This is a reversible state change (the fact is marked, not deleted), placing it in the Write category. However, since the description is truncated ('il n' is cut off), there is some uncertainty about whether it also deletes the fact, which would make it Destructive. Confidence is reduced accordingly.
From the tool's definition 'Marque un fait comme expiré' (marks a fact as expired) — the description is truncated but indicates the tool modifies the state of a fact by marking it as expired/invalidated
Attacks that exploit this kind of access
Marque un fait comme expiré (il n. It is categorised as a Write tool in the Memoraeu MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Memoraeu MCP server in PolicyLayer and add a rule for invalidate_fact: 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 Memoraeu. Nothing to install.
invalidate_fact 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 invalidate_fact 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 invalidate_fact. 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.
invalidate_fact is provided by the Memoraeu MCP server (pquattro/memoraeu-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 →