AI agents call memory_autoinject_test to retrieve information from Mementos without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This is a diagnostic/testing tool that queries memory matching without modifying state. It retrieves and displays information about what would happen in a hypothetical scenario, making it a Read operation with no destructive, write, execute, or financial implications. Low severity due to limited blast radius if misused—worst case is viewing internal memory matching logic.
From the tool's definition Tool description states it shows/tests what memories would be activated 'WITHOUT pushing' — explicitly read-only, no side effects. The purpose is to preview auto-inject behavior for testing/tuning parameters (min_similarity).
Attacks that exploit this kind of access
Test what memories would be activated by a given context WITHOUT pushing. Shows what the auto-inject pipeline would match — useful for tuning min_similarity. It is categorised as a Read tool in the Mementos MCP Server, which means it retrieves data without modifying state.
Register the Mementos MCP server in PolicyLayer and add a rule for memory_autoinject_test: 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 Mementos. Nothing to install.
memory_autoinject_test 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 memory_autoinject_test 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_autoinject_test. 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_autoinject_test is provided by the Mementos MCP server (@hasna/mementos). 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 →