AI agents call passoff_load to retrieve information from Passoff without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves and loads previously stored handoff context between AI agents. It has no side effects—it queries and returns data without creating, modifying, deleting, or executing code. The severity is low because loading context poses minimal risk; the main concern would be unauthorized access to sensitive context, but that is a data classification issue rather than a tool capability risk.
From the tool's definition Tool name is 'passoff_load' and description states 'Load a previously created handoff', indicating retrieval of existing data without modification. The verb 'load' and context of 'pick up where another AI left off' confirm read-only data retrieval.
Attacks that exploit this kind of access
Load a previously created handoff. Use when the user asks to load a handoff, pick up where another AI left off, or starts a session by referencing prior context. Pass. It is categorised as a Read tool in the Passoff MCP Server, which means it retrieves data without modifying state.
Register the Passoff MCP server in PolicyLayer and add a rule for passoff_load: 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 Passoff. Nothing to install.
passoff_load 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 passoff_load 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 passoff_load. 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.
passoff_load is provided by the Passoff MCP server (themrgu/ai-agent-context-passoff). 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 →