Several AI tools focus on board placement, watching, and state transitions rather than on the basic creation or editing of records. These are useful when the assistant needs to move work through an existing process rather than author the data from scratch.

These tools are useful after an agent already knows which project or task it is working with and needs to move that work through an existing process. That might mean putting an item on a board, changing who is watching a project, or updating the state of work as progress happens.

• Board tools can look up boards, inspect cards, add a project or task to a board, move a card within a board, and remove a card without deleting the underlying project or task.
• Watching tools operate on the current person. For example, SetProjectWatch starts or stops the current person watching a project rather than changing another person's watch state.
• State-change tools can set a project or task to Active, Completed, or Cancelled. On projects, completion can optionally cascade to descendant work; cancellation does not automatically cascade.
• State changes contribute to the same history and notification patterns the rest of the plugin uses, so AI-driven updates still participate in project history and configured notifications.

This group is less about authoring new records and more about keeping existing work in the right place and state. In other words, it is where an assistant behaves more like an operator: adjusting workflow position, maintaining participation, and reflecting the current status of the work.