AI agents call collectMemory to retrieve information from Miadi without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves memory values without modifying, deleting, or executing arbitrary operations. The sibling tools (gatherMemoryValues, getMemory, scanKeys) further confirm this is a memory query/read operation. No side effects or irreversible actions are described. Classified as Read with low severity and high confidence.
From the tool's definition Tool name 'collectMemory' and description 'Collect specific memory values by keys' indicate data retrieval. The verb 'collect' and operation 'by keys' are consistent with querying/fetching existing memory data without modification.
Attacks that exploit this kind of access
Collect specific memory values by keys. It is categorised as a Read tool in the Miadi MCP Server, which means it retrieves data without modifying state.
Register the Miadi MCP server in PolicyLayer and add a rule for collectMemory: 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 Miadi. Nothing to install.
collectMemory is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the collectMemory 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 collectMemory. 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.
collectMemory is provided by the Miadi MCP server (miadi-mcp-server). 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 →