Switch the active project context (affects which project new memories are saved to).
AI agents use switch_project to create or update resources in mcp-Agentmemory — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your mcp-Agentmemory environment.
switch_project is a state-mutating operation that changes the active project context. While it doesn't directly create or delete data, it configures the destination for future memory operations. This is a Write-category action because it modifies system state in a reversible way (the project context can be switched back).
From the tool's definition Tool changes which project new memories are saved to, modifying the system state that determines where subsequent write operations are directed.
Attacks that exploit this kind of access
Switch the active project context (affects which project new memories are saved to). It is categorised as a Write tool in the mcp-Agentmemory MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the mcp-Agentmemory MCP server in PolicyLayer and add a rule for switch_project: 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-Agentmemory. Nothing to install.
switch_project 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 switch_project 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 switch_project. 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.
switch_project is provided by the mcp-Agentmemory MCP server (obidel/agentmemory). 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 →