既存判断を更新(同一key維持)
AI agents use memory_supersede to create or update resources in MCP Memory System — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your MCP Memory System environment.
The tool updates an existing memory entry in-place while keeping the same key, which is a reversible modification of stored data. This is a Write operation. Severity is medium because overwriting decisions/rules that guide AI behavior across sessions could have meaningful downstream effects, but the operation appears reversible (prior values could be restored or re-saved).
From the tool's definition 既存判断を更新(同一key維持) — 'update existing decision (maintaining same key)'
Attacks that exploit this kind of access
既存判断を更新(同一key維持). It is categorised as a Write tool in the MCP Memory System MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the MCP Memory System MCP server in PolicyLayer and add a rule for memory_supersede: 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 MCP Memory System. Nothing to install.
memory_supersede 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 memory_supersede 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_supersede. 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_supersede is provided by the MCP Memory System MCP server (workerportfolio/mcp-memory-system). 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 →