SWARM TOOLS

33 tools from the Swarm MCP Server, categorised by risk level.

// ALL 33 SWARM TOOLS

READ 29 tools

Read beam beam Read classify classify Read decrypt decrypt Read encrypt encrypt Read filter filter Read get_type_definition Get a type definition by name, optionally resolving [references] to other types. Read guard guard Read hunt hunt Read inspect inspect Read iterate iterate Read list_governor_specs List all registered LLM-governed governors. Read list_pipelines List recent pipeline runs and their current status. Read list_sandbox_specs List all saved sandbox specs from ~/.claude/sandboxes/. Read list_type_registry List all registered types from ~/.claude/types/. Read map map Read map_reduce map_reduce Read par par Read pipeline pipeline Read pipeline_artifacts pipeline_artifacts Read pipeline_kill Kill a running pipeline and all its Docker containers. Read pipeline_status Return the current status of a running or completed pipeline. Read race race Read reduce reduce Read retry retry Read score_list score_list Read unwrap unwrap Read validate validate Read wrap Wrap a file or directory into the swarm ref system. Read wrap_project wrap_project

WRITE 2 tools

Write save_governor_spec save_governor_spec Write save_sandbox_spec save_sandbox_spec

EXECUTE 2 tools

Execute chain Run agents sequentially as a pipeline. Each stage receives the prior stage's output as context. Execute run run

// FAQ

How many tools does the Swarm MCP server have? +

The Swarm MCP server exposes 33 tools across 3 categories: Read, Write, Execute.

How do I enforce policies on Swarm tools? +

Route the Swarm server through the PolicyLayer gateway. Define allow, deny, or approval rules per tool in the dashboard; they are enforced on every call before it reaches the server.

What risk categories do Swarm tools fall into? +

Swarm tools are categorised as Read (29), Write (2), Execute (2). Each category has a recommended default policy.

Enforce policy on every Swarm tool call.

Start from Swarm, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.

CHECK YOUR STACK →

Instant setup, no code required.

43,000+ MCP servers and 220,000+ tools scanned and risk-classified.