Playbook plans and directives on the action board

What this page is

The Action Plans board is where HQ watches its own initiatives move. Every playbook, directive, recommendation, and experiment you have activated across the network shows up here as a single in-flight row, with a four-step lifecycle timeline: Drafted, Activated, Adoption check, Completed. It sits alongside three other kinds of network work in progress (franchisee interventions, consent requests, program audits and violations), but this article is about the one that starts as a decision you authored: the plan rows sourced from the network data.

Verinode does not run these playbooks for you. HQ writes the plan, activates it, and the network intelligence layer tracks where it stands and reflects that back as a timeline. Leadership decides what to launch and when to call it done; the board just keeps the "who launched what, and how far along is it" question answered without a status meeting.

Where to find it

Open Actions from the HQ sidebar, or go directly to hq.verinode.ai/actions. It loads to the Gantt view by default. A second view, Calendar, is available from the toggle in the top-right of the filter bar; a search box, a bucket dropdown, and a sort control sit to its left.

The board pulls the same underlying rows as the sibling Decisions page (/franchise/decisions), described in Decisions: the franchisor inbox. The difference is scope: Decisions shows every network row regardless of status, including plans still in draft; Actions narrows to only the rows currently actioned (an activated plan) or resolved (a completed one). A plan you have not yet activated, or one you paused or archived, will not appear on this board at all, it lives on Decisions instead.

Where these plans come from

Playbooks, directives, recommendations, and experiments are authored on Broadcast → Initiatives (/franchise/broadcast), not on the Actions page itself, using the + New decision form. When you author one, you pick one of four kinds:

• Playbook, a canonical multi-step playbook for the network
• Directive, a one-shot franchisor directive: do this, by this date
• Recommendation, a soft nudge with no enforcement
• Experiment, an A/B test run at the network level

Every new decision lands in draft status first. You move it to Active from its detail view when it's ready to deploy, at which point it starts appearing on this Actions board. See Broadcasting to your network for the authoring flow. This article picks up once a plan goes live.

The four-step timeline

Every plan row on the board carries the same four checkpoints, regardless of which of the four kinds it is:

1. Drafted: the day the plan was created. Always shown as complete, since the row only reaches this board after activation.
2. Activated: the day you moved the plan from draft (or paused) to active. Also always shown as complete on this board, since activation is what got the row here in the first place.
3. Adoption check: a checkpoint roughly two weeks after activation (day 14), meant as a natural point to look back and ask whether the network is actually adopting the plan. It shows complete only once the plan itself is marked completed; until then it sits as the plan's next open step.
4. Completed: the day you mark the plan completed. Shown as complete only after you have done that from the plan's detail view or via the Act control described below.

Verinode builds this timeline from the plan's own timestamps (created_at, activated_at, completed_at), not from a separate log of status changes, so the Adoption check date is a fixed two-week guideline rather than a check Verinode performs for you. If a plan runs its course faster than two weeks, the Adoption check checkpoint is timestamped just ahead of the completion date instead of at day 14, so the four steps stay in order on the timeline.

Reading the Gantt

The default view is a weekly Gantt spanning about four months (16 weeks, starting one week back from today), the same construction-style timeline used on the operator side of Verinode.

Header. The top-left corner reads Active plans and a count, "N plans underway," where N is every row currently on the board (plans, interventions, consent requests, and program items combined). Above the weekly grid, a second header row groups columns by month. The current week is highlighted and labeled This week; every other column reads Week of plus its start date.

Each row. The left-hand label column shows the plan's title. Decision plan rows do not carry an entity name or a dollar figure (unlike an intervention or a program violation, which show a franchisee or program), so the row is just the title plus its progress: X/N steps and a thin copper progress bar showing how many of the four checkpoints are marked complete.

The bar itself. To the right of the label, a segmented horizontal bar runs from the plan's activation date to its projected completion date, with a vertical copper line marking today. Each of the four steps is a separate segment, colored by its own status:

• Deere Green: that step is complete
• Ember Red: that step is overdue
• Copper (full): due today
• Copper (light): due tomorrow
• Copper (faint): upcoming, not yet due
• Muted gray: not started

A small white checkmark appears inside a completed segment if it's wide enough to show one. Hovering any segment opens a tooltip with the step's name, its status (Completed, Overdue, Due today, Due tomorrow, or Upcoming), and its due date, plus two buttons: Mark complete / Undo complete, and Open plan. Clicking a segment, the plan's title, or anywhere on its row opens the full decision workspace for that plan.

A legend along the bottom of the Gantt card repeats the five status colors, with a reminder: "Hover a step for details, click the bar or any step to open the plan."

Filtering and sorting

The dropdown next to the search box buckets every row on the board by the urgency of its next open step, not by raw status:

• All: every in-flight and completed row
• Planning: the agent is actively generating a plan (this bucket never applies to decision plans; it's populated only by AI-generated operator plans elsewhere on the platform)
• Ready to start: a plan has been saved but not yet started (also not applicable to decision plans, which are already running by the time they reach this board)
• Overdue: the next open step is past its due date
• Due today / Due tomorrow: self-explanatory
• Later: nothing due yet
• Completed: every step is marked done

For a decision plan, that means an active, not-yet-completed plan will land in Overdue, Due today, Due tomorrow, or Later depending on where its Adoption check step sits relative to today; once you mark it completed, it moves to the Completed bucket.

The search box filters by title (and, for other row kinds, by franchisee or program name). The sort control offers Newest first (default) or Biggest impact; decision plans do not carry a dollar impact figure, so a sort by impact naturally sorts them after any row that does.

Switching to Calendar view lays the same rows out on a month grid instead, one dot per day a step is due, built from the same underlying evidence.agent_plan timeline as the Gantt. If nothing is on the board yet, the calendar reads: "No plans on the calendar yet. Action plans appear here as soon as you click Act on a decision. Each plan step lands on the day it is due."

Acting on a plan from this board

Opening a plan (clicking its row or a step) drops you into the decision workspace, a glass overlay showing the plan's title and the detail text you wrote when you authored it, with arrows (or swipe, on touch) to step to the next or previous plan in your current filter. From there, or from the plan's row on Decisions, HQ admins can move a plan through its lifecycle:

• Act advances the plan one stage: a draft or paused plan becomes active; an already-active plan becomes completed.
• Park (available only while a plan is active) pauses it, which also removes it from this board, since paused plans only show on Decisions.
• Ignore archives the plan at any stage. Archived plans also drop off both boards.
• Discuss opens the network AI agent panel so you can talk through the plan without changing its status.

Only group admins can take these actions; other HQ roles can view the board and open a plan to read it, but Act, Park, and Ignore are not available to them.

Empty states

If there is nothing on the board at all for your current filter, the Gantt reads plainly rather than showing a blank chart. The exact wording depends on what you have filtered to:

• No plans match the filter at all: "Nothing matches this filter."
• Filtered to Overdue: "Nothing overdue. Steps past their due date land here so they don't slip."
• Filtered to Due today: "Nothing due today. Steps scheduled for today land here as soon as a plan crosses their day_offset."
• Filtered to Due tomorrow: "Nothing due tomorrow yet. Tomorrow's step list builds as the day rolls over."
• Filtered to Later: "Nothing parked for later. Plans you don't act on today land here so you can pick them up next week."
• Filtered to Completed: "No completed plans yet. Plans you resolve land here so you can run a retrospective."

A board with zero playbooks or directives activated yet is not a broken page, it means nothing has been moved from draft to active. Author one on Broadcast → Initiatives, activate it, and its row appears here on the next load.

The privacy boundary

Decision plans are HQ's own content, authored at the network level, so there is nothing to anonymize on this particular row type: the title and detail text are what you wrote. What stays boundaried is everything downstream of the plan. HQ sees aggregate adoption signal (how many franchisees the plan reached, in what state) on the Decisions playbook scoreboard, never a single franchisee's underlying business data. The other row kinds that share this board, franchisee interventions and consent requests, anonymize the franchisee's identity by default unless your network is set up as a single legal entity; HQ never sees an individual franchisee's financials or operations because a plan targeting them showed up on this board.

Related reading

• Decisions: the franchisor inbox
• Broadcasting to your network
• HQ programs and compliance
• HQ compliance and program violations
• Network health