Agent Branch

An agent branch is a separate working version of a project where AI-made changes can be reviewed before they affect the main site.

An agent branch protects the live project by isolating AI edits. Teams can preview the result, compare changes, request revisions, and merge only when the work is ready.

Related terms

Related terms

  • Hooks

    AI

    Hooks are automation points tied to lifecycle events that let teams enforce checks, trigger scripts, or modify behavior around agent actions.

  • Approval Modes

    AI

    Approval Modes are policy settings that control which operations an agent can run automatically and which require explicit human approval.

  • Branch Preview

    Publishing

    A branch preview is a shareable version of work on a branch, used to review changes before merging or publishing them live.

    In collaborative Framer workflows, branch preview helps teams move faster without losing control of structure, content, performance, or editable design details.

  • Merge to Main

    Publishing

    Merge to main is the action of bringing approved branch changes back into the primary project version before shipping them.

    In collaborative Framer workflows, merge to main helps teams move faster without losing control of structure, content, performance, or editable design details.

  • Agent Audit Trail

    AI

    An agent audit trail is a record of AI-assisted changes, decisions, and updates that helps teams review what happened in a project.

    An agent audit trail makes AI work accountable. It can include changed content, branch activity, review notes, or publishing history so teams can understand what was edited and roll back or refine work when needed.

  • Agent QA

    AI

    Agent QA is the use of an AI agent to find and help fix issues such as broken links, duplicate content, missing metadata, or inconsistent layout.

    Agent QA supports quality assurance by scanning project content and structure for problems. It is most useful when paired with human review, especially before merging a branch or publishing a site.

  • Conversational Editing

    AI

    Conversational editing is the process of changing a website through natural-language feedback instead of direct manual edits alone.

    Conversational editing lets a user describe what should change and have an agent update the project. It works best when each request names the target clearly and the agent keeps the result editable for follow-up refinements.

  • Agent-Orchestrated Workflow

    AI

    An agent-orchestrated workflow coordinates multiple AI actions, tools, or agents to complete a larger project task.

    Agent-orchestrated workflows are useful for multi-step site work such as scanning pages, creating CMS entries, adding links, reviewing changes, and preparing a branch for approval. The orchestration keeps each action aligned with the overall goal.