Palma

Palma is an enterprise platform designed to manage, govern, and audit AI agents that use the Model Context Protocol (MCP). It acts as a control layer between your AI agents—whether human-assisted tools in IDEs like VSCode or fully autonomous systems—and your core business applications like databases, CRMs, and project management tools. The primary function is to ensure that every tool call an agent makes is secure, compliant, and cost-effective.

Palma works by implementing the MCP open standard, creating a centralized catalog of connected tools. It allows administrators to set granular access policies, require human approval for sensitive actions, and prune or scope the context sent to agents for accuracy. What sets Palma apart is its focus on the operational side of AI agent deployment, providing full visibility into who (or what agent) is using which tools, how much it costs, and whether the actions comply with company rules. It is built to support agentic workflows from the ground up.

This tool is for organizations that are scaling their use of AI agents and need to maintain control. Enterprise IT teams, AI operations managers, and security officers benefit most. Use cases include giving a development team secure access to database tools via an AI assistant while tracking costs, or ensuring an autonomous agent fleet handling customer support tickets only accesses approved data and escalates certain requests to a human. Palma provides the oversight needed to trust AI agents with critical business systems.