agent-filesystem

Use when the user wants to migrate Codex state in ~/.codex into Agent Filesystem and mount the same shared Codex memory/settings across multiple computers. Recommends a .afsignore before migration and defaults to excluding worktrees, caches, logs, and temporary files.

Use when agents need persistent shared storage, when saving or restoring workspace state, or when coordinating file access across multiple agents and machines. Creates Redis-backed workspaces, checkpoints and restores agent state, mounts shared filesystems locally, searches workspace contents, and forks workspaces for parallel work.

Manage Agent Filesystem through the control-plane MCP server. Use when the `agent-filesystem` MCP server is available, or when the user asks Codex to list, create, fork, checkpoint, delete, or connect AFS workspaces using a control-plane token.

Apply the team's shared coding rules before writing, modifying, or reviewing code. Trigger when the user asks to add a feature, refactor, fix a bug, write tests, or review a change. Reads rules live from the team-rules MCP workspace at /rules/*.md.

Read and update the team's shared PRD in the team-prd AFS workspace. Trigger when the user asks about current/inflight/done work, wants to start a new work item, mark something complete, or add an open question to the team's spec. Enforces per-dev subdir isolation to avoid write conflicts between multiple developers' agents.

Use when agents need persistent shared storage, when saving or restoring workspace state, or when coordinating file access across multiple agents and machines. Creates Redis-backed workspaces, checkpoints and restores agent state, mounts shared filesystems locally, searches workspace contents, and forks workspaces for parallel work.