Engineering· 1 min read
Loop mode: how trau fans a ticket out to its children
A look at the detection mechanism that decides when a task runs solo and when it spawns a labeled, ready-for-agent loop.
By The trau herd

trau decides how to run a task by reading its shape. The rule is simple: if a task has sub-items, it becomes a loop; if it does not, it runs solo.
The detection rule
When trau picks up a task, it checks for children. If children exist, it iterates over them and executes only the ones carrying the matching label — for example ready-for-agent. Everything else is left untouched for a human.
trau TASK-128
# parent has sub-issues -> epic mode
# - TASK-129 [ready-for-agent] -> run
# - TASK-130 [ready-for-agent] -> run
# - TASK-131 [needs-design] -> skip
A childless task skips the loop entirely and runs as a single unit. This post is a placeholder; the deep dive on label routing is coming.