backlog-burndown re-dispatches a worker onto an existing branch at three sites, and no procedure covers attaching a worktree to a branch that already exists.
The three sites: "Dispatching a Worker"'s local-strict-review bullet ("A clean tree is retired and the group re-dispatched", where "the commits the branch already carries are what the re-dispatched worker continues from"), "Raising a Blocked Question"'s "Resuming retires that worktree first, then dispatches a fresh worker", and "The Promotion Boundary" step 5's "dispatches a fresh worker on that same branch".
All three ground worktree creation in "drive-pr step 1 and repo-worktree's task-start mandate", and repo-worktree "Creating a Worktree" gives exactly one form:
git -C ~/repos/<Repo> worktree add ~/repos/worktrees/<Repo>-<task-slug> -b <task-branch> origin/develop
-b fails on a branch that already exists, and if it succeeded it would discard the commits the first site says the worker continues from. The skill needs to name the attach-an-existing-branch form, or say that a re-dispatch brief overrides the -b recipe.
Related, in the same bullet: "A clean tree is retired and the group re-dispatched to a seat that can dispatch" has no reachable second seat. The only other seat this skill names is the orchestrator, which "opens no feature branch and fixes no issue itself", and a replacement worker is dispatched by the same mechanism that just proved it cannot dispatch. Nothing says what would make a second seat different, so in practice only the "no seat that can dispatch exists" arm is reachable.
Raised by the local-strict-review carried-content passes on the pull request that added this skill, and deferred there under that skill's own review-round budget.
backlog-burndownre-dispatches a worker onto an existing branch at three sites, and no procedure covers attaching a worktree to a branch that already exists.The three sites: "Dispatching a Worker"'s
local-strict-reviewbullet ("A clean tree is retired and the group re-dispatched", where "the commits the branch already carries are what the re-dispatched worker continues from"), "Raising a Blocked Question"'s "Resuming retires that worktree first, then dispatches a fresh worker", and "The Promotion Boundary" step 5's "dispatches a fresh worker on that same branch".All three ground worktree creation in "
drive-prstep 1 andrepo-worktree's task-start mandate", andrepo-worktree"Creating a Worktree" gives exactly one form:-bfails on a branch that already exists, and if it succeeded it would discard the commits the first site says the worker continues from. The skill needs to name the attach-an-existing-branch form, or say that a re-dispatch brief overrides the-brecipe.Related, in the same bullet: "A clean tree is retired and the group re-dispatched to a seat that can dispatch" has no reachable second seat. The only other seat this skill names is the orchestrator, which "opens no feature branch and fixes no issue itself", and a replacement worker is dispatched by the same mechanism that just proved it cannot dispatch. Nothing says what would make a second seat different, so in practice only the "no seat that can dispatch exists" arm is reachable.
Raised by the local-strict-review carried-content passes on the pull request that added this skill, and deferred there under that skill's own review-round budget.