Process
Start here
A flow is a list of steps. Each step points at the step that comes next.
Use an action for something you do, a decision when the path splits on a yes/no,
a parallel step when several things happen at once, and an end to close a path.
Swimlanes are columns: put a separate workstream in its own lane.
Swimlanes
Steps
Runbook header
Steps
0m
Now
Timeline
0m
Events
What this is
flow-state turns a process into three views of the same thing: a list of steps you edit,
a flowchart you can print or hand over, and a runbook with the detail each step needs.
Build the steps on the Build tab, read the chart on the Flowchart tab, fill in the detail
on the Runbook tab, and save the config to carry all of it.
Everything runs client side. There is no network call and no analytics. The config file is plain JSON, so it diffs and reviews like any other artefact.
The Run tab is the other half. Import a config, start a run, and the tool walks the flow with you: it shows only the steps that are live right now, you pick the outcome at each decision, and every choice, note and skip is timestamped into a record you can export or print. Off-script actions get logged too rather than being lost. Imported processes, run records and incident timelines are kept in this browser separately from the flow you are building, so resetting the builder does not touch them.
The Timeline tab is for the incidents that do not follow a process at all: log events with a time and a source, in whatever order you learn them, and it keeps them in sequence and exports the same way everything else does.
Everything runs client side. There is no network call and no analytics. The config file is plain JSON, so it diffs and reviews like any other artefact.
The Run tab is the other half. Import a config, start a run, and the tool walks the flow with you: it shows only the steps that are live right now, you pick the outcome at each decision, and every choice, note and skip is timestamped into a record you can export or print. Off-script actions get logged too rather than being lost. Imported processes, run records and incident timelines are kept in this browser separately from the flow you are building, so resetting the builder does not touch them.
The Timeline tab is for the incidents that do not follow a process at all: log events with a time and a source, in whatever order you learn them, and it keeps them in sequence and exports the same way everything else does.