Browse all 9 articles in this open-source series. Nine practical lessons on Git worktree coordination for AI development, with pinned sources, seven teaching-model cases and explicit runtime limits.
Featured starting point
Understand how Worktrunk coordinates branches, directories and lifecycle commands, and why a separate worktree is not an agent sandbox.
Suggested learning path
01 → 09
Understand how Worktrunk coordinates branches, directories and lifecycle commands, and why a separate worktree is not an agent sandbox.
Learn the smallest worktree exercise, separate successful creation from shell navigation, and avoid accidental hooks or cleanup.
Separate package installation from shell integration and repository hooks, with a versioned rollout and a narrow rollback plan.
Trace control from an operation through configured commands, approval and pre/post hooks, including startup dependencies and merge behavior.
Walk through approve_command_batch and approval storage to understand why accepting a command is different from remembering that acceptance.
Build a practical ledger for checkout preparation, builds, model calls and review time before deciding whether more agents improve throughput.
Review actual command authority, optional model data flow and recovery limits before automating merge or cleanup.
Compare manual worktree management, Worktrunk lifecycle automation and agent orchestration without treating them as interchangeable products.
Design a small audit report for branches, configuration and unresolved risks without automatically approving hooks, merging code or deleting worktrees.
