// MCP TOOL REFERENCE

LOOPIN TOOLS

6 tools from the Loopin MCP Server, categorised by risk level.

READ TOOLS

decide_interrupt Submit a human decision on an interrupt request (approve or reject). get_interrupt_analytics Get interrupt usage analytics for a user — total count, approval rate, average response time, and top action types. get_interrupt_status Poll the status of an interrupt request. Returns current status and the decision if one has been made. list_pending_interrupts List all pending interrupt requests for a user. Use to see what decisions are waiting for human review.

WRITE TOOLS

create_interrupt Create a new human-in-the-loop interrupt request. Use when your agent hits a decision point requiring human approval before proceeding. 2/5

DESTRUCTIVE TOOLS

cancel_interrupt Cancel a pending interrupt request. Use when the agent no longer needs the decision. 4/5

// FAQ

How many tools does the Loopin MCP server have? +

The Loopin MCP server exposes 6 tools across 3 categories: Read, Write, Destructive.

How do I enforce policies on Loopin tools? +

Use Intercept, the open-source MCP proxy. Write YAML rules for each tool — rate limits, argument validation, or deny rules — then run Intercept in front of the Loopin server.

What risk categories do Loopin tools fall into? +

Loopin tools are categorised as Read (4), Write (1), Destructive (1). Each category has a recommended default policy.

Enforce policies on Loopin

Open source. One binary. Zero dependencies.

npx -y @policylayer/intercept

github.com/policylayer/intercept →