Skip to main content

​
Overview

The pipeline is the same table seen as a kanban board: you define the steps, and every row becomes a card sitting in one of them. It’s how you work a researched list rather than just read it, moving each row through the stages you define as things progress. Switch to it with the pipeline icon in the table header. Boards are free and instant: configuring one, moving cards, and reading a board never cost credits.

​
Setting up a board

The first time you open the pipeline on a table, Hizura asks you to set up its steps. Pick steps that fit what the table is actually for. A sales table might use To contact → Contacted → Moving forward → Success; a hiring table or a fundraising list wants something else entirely. There’s no fixed funnel. You also choose which columns appear on each card. The entity name is always the card title, so you never pick it as a card column.
Setting up or changing a board needs edit access. If you’re a read-only collaborator on a table that has no board yet, the pipeline toggle stays disabled — ask the table’s owner to set one up.

​
Working the board

  • Move a card by dragging it to another step
  • Reorder steps by dragging their headers
  • Rename a step by clicking its title and typing — a renamed step keeps its cards
  • Delete a step once it’s empty
  • Collapse a step by double-clicking its header
Each step header shows a live count of the cards in it.

​
Card density

The toggle beside the view switch controls how much each card shows: Use the /pipelineViewCols command to change which columns appear on cards.

​
What’s shared and what isn’t

This distinction matters on a shared table: So moving a card is a change your teammates see, but switching yourself to board view doesn’t move anyone else’s screen.

​
Driving the board from an MCP client

Boards are fully controllable over MCP:
  • get_pipeline — read the steps, the cards in each, and the column ids valid for cards
  • configure_pipeline — create or rewrite the board’s steps
  • move_cards — move many rows between steps in a single call
Sorting a whole table into steps in one call is the small version of this. The real one is that the board becomes the surface your agents work through. get_pipeline hands an agent exactly what it needs to act: who the accounts are, what’s known about each, and what stage every one of them is at. Give it the rest of its toolbelt — mail, calendar, whatever else it can reach — and it can work through a stage, doing whatever that stage calls for, then move each card to reflect what actually happened. Your steps are what assign the work — an account’s stage says what it’s waiting for — and moving it on is how the agent reports back. What you’re left with is a board that records what your agents did, in the same place your team already looks. You read it and correct it rather than operating it — and because it’s one shared object, a rep dragging a card and an agent moving one are the same action on the same thing.
configure_pipeline takes the complete ordered list of steps. Any step you leave out is deleted and its cards fall back to the first step — so read the board with get_pipeline first and pass every existing step plus your new one.