The Gantt timeline: reading plans on the 30-day ruler

What the Gantt timeline is

Action Plans is the network's view of the work already in flight: plans activated, interventions being worked, consent requests approved, program actions moving. Where Decisions is the franchisor's inbox (everything pending a call), Action Plans is the execution layer underneath it, the initiatives HQ has already committed to and is now driving to completion. The Gantt timeline is the default way that work renders: one horizontal row per initiative, its synthesized steps drawn as chunky, color-coded blocks along a day ruler.

Verinode is not a job-management tool or a project tracker. Every initiative you see here is one HQ itself is running (a decision plan it activated, an at-risk-member intervention, a governance ask, a program action) not a franchisee's private job schedule. The timeline is HQ's own operating rhythm across the network, never a window into an individual member's business.

Where to find it

Open Action Plans from the HQ sidebar (it sits in the top group, right below Decisions: Feed, Decisions, Action Plans, Playbooks, Broadcast, Vault). The URL is hq.verinode.ai/actions. The page opens straight into the Gantt view, that is the default for this surface.

A view toggle in the top-right of the filter bar switches between two reads of the same filtered row set:

• Gantt (grid icon, three stacked bars), the timeline described in this article.
• Calendar (calendar icon), a day-by-day read of the same steps.

Gallery view, the tile-grid layout used on Decisions, is not available here. A single flat decision tile is the wrong read for a multi-step, time-bounded plan, so Action Plans only ever shows Gantt or Calendar.

What counts as a row here

Action Plans pulls from the same four network sources as Decisions: activated decision plans, the at-risk-member intervention queue, HQ governance consent requests, and program audits and violations. It pre-filters to only the initiatives currently actioned or resolved, anything still pending, parked, or dismissed belongs on the Decisions board, not here. The header label under "Active plans" (top-left of the Gantt) reads the exact count of rows on screen, e.g. "6 plans underway."

A row only appears in the Gantt once it has a synthesized plan with at least one step, or an agent run actively generating one. Rows without steps stay on Decisions and never surface here.

Reading the ruler

The chart is a 16-week window (about four months), one column per week, each column 92 pixels wide so a step label has room to breathe. Two header rows orient you:

• Month row, the month name and year, spanning however many week columns fall inside it.
• Week row, each column's Monday date. The current week is highlighted (a light copper wash + a copper bar across the top of the cell) and labeled "This week" instead of "Week of." Every other column reads "Week of" plus its Monday's date.

A thin copper vertical line runs down the chart at today's exact position, inside the current week if you look closely, this is the "you are here" marker the rest of the ruler is read against.

The window starts one week before today (so the immediate past stays visible for context) and one Monday of the current week is always inside the visible range. If you have more rows than fit on screen, the chart scrolls vertically inside its own card, and if the window's 16 weeks run wider than your screen, it scrolls horizontally, both stay inside the card, the page itself does not scroll while you're in Gantt view.

Anatomy of one row

Each row has two parts: a fixed label cell on the left (280 pixels wide, stays pinned in place as you scroll the ruler sideways) and the timeline track on the right.

The label cell

• Title, the initiative's name, e.g. "Reduce cycle time on water mitigation claims" or an intervention's flag reason.
• Member name, shown underneath when the row is tied to a specific franchisee (an intervention or a consent request). Consistent with the network-wide privacy boundary, this is the anonymized location label HQ sees for that member, never the franchisee's raw internal name unless your group is configured as a single legal entity.
• Impact dollar figure, shown when the initiative carries an estimated network-level impact, formatted as $XXk or $X.XM with a period suffix (/yr, /mo, /job, or "one-time").
• Progress line, one of three states depending on where the plan sits:

- Planning, a pulsing copper pill reading "Planning." This means an agent is actively generating the step-by-step plan right now, there is nothing to show on the track yet because the steps don't exist. - Ready to start, the step count (e.g. "5 steps") plus an Execute Plan button. The plan has been generated but nobody has kicked it off yet, there's no start date, so there's nothing to anchor the timeline bar to. Clicking Execute Plan starts the plan immediately from the row, no need to open the full workspace first. - Started, an "X/N steps" counter and a thin copper progress bar showing what fraction of the plan's steps are checked off. This is the normal state for a plan in motion.

Clicking anywhere in the label cell (except the Execute Plan button) opens the workspace slider for that initiative, same as clicking a tile on Decisions.

The timeline track

For a Planning or Ready to start row, the track is just the empty week-column grid, no bar, nothing to click. Once a plan is running, the track shows a single chunky rounded bar spanning from the plan's start date to its last step's due date, divided into segments, one per step, separated by a thin "grout line" so it reads like poured, jointed concrete rather than a smooth gradient.

Each segment is colored by that step's status:

| Status | Meaning | Color | |---|---|---| | Completed | Step is checked off | Deere Green (Expand signal color) | | Overdue | Due date has passed and it's not done | Ember Red (Analyse signal color) | | Due today | Due date is today | Full copper | | Due tomorrow | Due date is tomorrow | Copper at 80% | | Upcoming | Due date is further out | Copper at 40% |

A completed segment shows a white checkmark centered in it, if the segment is wide enough to fit one. A legend along the bottom of the card (desktop only) spells out all five colors plus the instruction: "Hover a step for details, click the bar or any step to open the plan."

Working a step: hover and click

Hover any segment to open a small glass tooltip above it. The tooltip shows:

• The step number and its current status (e.g. "Step 3 · Overdue").
• The step's title.
• Its due date (e.g. "Due Wed Apr 10").
• Two buttons: Mark complete (or Undo complete if it's already done), and Open plan.

Mark complete flips the step's status immediately in the timeline (the segment recolors to green and the label-cell progress counter ticks up) while the change saves in the background. If the save fails, the step reverts and the tooltip stays available to retry.

Click any segment, or the bar itself, or the row's label, and the workspace slider opens for that initiative. Inside the slider you can inspect the full plan, every step with its rationale, and (for the row kinds HQ is allowed to act on directly) the same Act, Park, Ignore, and Discuss controls exposed on the Decisions board. Which of those controls appear depends on the initiative's kind and current status:

• Decision plans: draft or paused plans can be activated; active plans can be marked complete; any state can be archived.
• Interventions: open or in-progress interventions can be advanced through their lifecycle; any state can be marked no-action.
• Consent requests: HQ can only withdraw a request, franchisees are the ones who approve or decline it.
• Program audits: submitted or disputed audits can be approved; submitted audits can be disputed.
• Program violations: violations can be escalated, resolved, or waived at any open state.

Every row also carries a Discuss action that opens the AI agent panel with that specific initiative in context, useful for asking the agent to draft an update, summarize where a plan stands, or suggest a next move without leaving the timeline.

Row ordering

Rows are not sorted alphabetically, they're sorted by urgency, so the initiatives needing your attention float to the top automatically:

1. Planning rows first (an agent is actively working, worth a glance).
2. Ready to start rows next (a plan is waiting on your Execute Plan click).
3. Among started plans, any row with an overdue step outranks one without.
4. Among the remainder, a row with a step due today outranks one without.
5. Everything else sorts by how soon its next incomplete step is due, soonest first.

This means the top of your Gantt is always the plans that most need eyes on them right now, not the plans that happen to have alphabetically early titles.

Filtering what you see

The filter bar above the Gantt narrows the row set before it hits the timeline:

• Search, a free-text box matching the initiative's title or its member name.
• Status bucket, a dropdown that replaces the Decisions status filter with plan-step urgency buckets specific to this surface: All, Planning, Ready to start, Overdue, Due today, Due tomorrow, Later, Completed. Each option's count reflects how many actioned or resolved initiatives currently fall in that bucket.
• Sort, Newest first or Biggest impact, this reorders the underlying row set before the urgency-based Gantt ordering above is applied within it.
• Select, switches the page into a flat checkbox list so you can cancel several in-flight plans in one pass, useful for a batch clean-up rather than closing plans out one at a time. A sticky bar at the bottom of the screen shows how many you've picked and lets you cancel them together; canceled plans close out as not completed and drop off Action Plans.

Domain filtering (the business-area dropdown used on Decisions) is intentionally not present here, plan-step urgency is the single axis that matters on this surface.

Empty states

If no initiative in your current filter has a synthesized plan yet, the card shows a dashed-border box instead of the ruler. The exact wording depends on why it's empty:

• No filter narrowed the list, nothing has a plan at all: "Nothing matches this filter."
• General (non-plan-bucket) empty: "You haven't started a plan on any of these yet. Open one in gallery to start the plan."
• Later bucket empty: "Nothing parked for later. Plans you don't act on today land here so you can pick them up next week."
• Today bucket empty: "Nothing due today. Steps scheduled for today land here as soon as a plan crosses their day_offset."
• Tomorrow bucket empty: "Nothing due tomorrow yet. Tomorrow's step list builds as the day rolls over."
• Overdue bucket empty: "Nothing overdue. Steps past their due date land here so they don't slip."
• Planning bucket empty: "No plans in the planning beat. New decisions you've accepted but not yet started will land here."
• Ready bucket empty: "Nothing ready to start. Plans whose first step has come due will land here."
• Completed bucket empty: "No completed plans yet. Plans you resolve land here so you can run a retrospective."

Best-practice example

Open Action Plans first thing in the morning with the status bucket set to All. Anything with a pulsing "Planning" pill is an agent mid-draft, worth a quick check-in later. Anything with an "Execute Plan" button is a finished plan waiting on you, click Execute Plan directly from the row rather than opening the workspace first, it starts the same instant. Then scan down the ordered list: rows with a red (overdue) segment come first by design, hover that segment, read the step and its due date, and either mark it complete if the work already happened or click Open plan to update the record. Once the overdue and due-today rows are clear, the rest of the board is genuinely "later," and you can move on to Decisions to see what new initiatives need a call.

Related reading

• HQ overview, how the HQ workspace is organized.
• The decision workspace, the slider that opens when you click a plan.
• Network health, where at-risk-member interventions originate.
• HQ programs, the source of program audits and violations that surface here.
• HQ compliance, the governance side of consent requests.

Data sources