AI agents use add_task to create or update resources in Memex — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Memex environment.
This tool creates or modifies data in the memex memory system by adding a new task. It falls under Write rather than Execute because it stores data persistently rather than triggering external operations. Severity is medium because task creation in an agentic memory system could cause the agent to pursue unintended goals or accumulate false commitments, but the effect is reversible.
From the tool's definition Tool name 'add_task' indicates creation of a task record in persistent memory; description is empty but the tool is part of a memory system (memex) that manages persistent data structures
Attacks that exploit this kind of access
add_task. It is categorised as a Write tool in the Memex MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Memex MCP server in PolicyLayer and add a rule for add_task: 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 Memex. Nothing to install.
add_task 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 add_task 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 add_task. 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.
add_task is provided by the Memex MCP server (queflyhq/memex). 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 →