What it does
policy-gate is an MCP server from fieldproofhq that evaluates actions an AI agent intends to take and returns one of three verdicts: allow, require_approval or deny. The verdict is produced before the action happens, so the agent can be checked against a rule set rather than acting unchecked.
It fits developers who build or run AI agents and want a fixed decision point between an agent's request and its execution, for example to hold certain actions for human approval or block them outright.
The publisher describes the verdicts as deterministic, meaning the same input is intended to produce the same verdict. The source material does not describe how rules are defined or how the server is installed and configured beyond its listing as an MCP server.
Safety
Our team has not reviewed this resource. Only automated rules checked it; resources with high-risk signals are not published. These are signals, not a guarantee that installing it is safe. Try it in a sandbox project first.
- Runs code or changes an assistant's behaviour
Sources
- Official MCP Registry first seen 13 Nov 2026 · last seen 13 Nov 2026 · its page on the source
dev.workers.3labsio.policy-gate/policy-gate
Deterministic allow/require_approval/deny verdicts for agent actions, before they happen.
v0.2.0As shown on registry.modelcontextprotocol.io when we last read it