About Golf
Golf is a security and governance platform for AI agents and the Model Context Protocol connections they use. Its premise is that engineering and business teams are connecting tools such as Cursor, Claude Code and GitHub Copilot to production data through MCP servers, and those connections bypass existing SIEM, data loss prevention and identity controls. Golf sits at the MCP layer, not the model layer, so agents keep using their own models while Golf governs where they connect.
The product works in three parts. Discovery finds every AI agent, MCP server and data connection in an environment, including ones nobody knew about. Enforcement applies granular policies per tool, team and data source, blocking exposure of personal data, credential leaks and unauthorized access, with instant rollback. Audit keeps a 90 day trail of prompts, actions and data access, mapped to frameworks such as SOC 2, ISO 27001, NIST AI RMF and FINRA, with evidence export. It is aimed at security, IT and governance teams in organizations that are adopting coding agents and enterprise AI assistants. The vendor lists SOC 2 Type II and describes itself as a Y Combinator company.
Golf does not publish pricing. The site directs visitors to book a call or request a demo, and it also offers a free Agentic Governance Playbook.