Agent Infrastructure
How CR Base provides the infrastructure agents run on.
Overview
CR Base provides the infrastructure that all agents run on.
Persistent Memory
Agents carry context across sessions. Brand rules, creative patterns, campaign history, and compliance decisions compound over time. Agents do not start from zero on every project. The more campaigns run through the system, the more context agents have to work with.
Governed Execution
Every agent operates inside the rules set by the platform. The Primary Agent has final authority. No agent can bypass compliance checks or brand rules. Agent permissions are scoped at creation time and can never exceed the permissions of the user who created them.
Audit Logging
Every agent action is logged. The log includes:
- What the agent ran
- What data it accessed
- What it produced
- What rules it followed
- Duration and credit cost
Agent execution history is exportable for compliance and legal review. Logs are stored in the evidence vault as immutable records.
MCP Compatibility
Agent infrastructure supports MCP (Model Context Protocol) for connecting agents to external systems and tools within governed boundaries. MCP connections are subject to the same permission scoping and audit logging as all other agent activity.