The triage log: restoring auto-archived decisions

What the triage log is

An active operator can accumulate hundreds of open decisions in a few weeks. Left unchecked, that turns Decisions and the Feed into a wall of noise. Verinode keeps both usable by quietly archiving decisions that have gone stale, expired, or been superseded, on a schedule, without you touching anything.

The triage log is the reversible audit of that housekeeping. It lists every decision the system, not you, archived in the last 60 days, tells you which rule or judgment call archived it and why, and lets you bring any one of them back with a single click. Nothing here is permanent. If Verinode's judgment was wrong for a given case, restoring it takes seconds.

This is strictly about system-initiated archiving. A decision you dismissed yourself, on purpose, from a decision card, never appears in this log. It is filtered out from the start: the triage log only ever shows rows the system flagged with its own archive record, so what you see here is exclusively "Verinode did this," never "I did this."

Where to find it

Open Decisions from the sidebar. In the top right of the page, next to Workload, click Triage Log. It opens as a glass overlay on top of the Decisions page, so your place in the log stays intact and you're one click from being back where you were. (A standalone page at /decisions/triage-log shows the identical content if you land on it directly, and the mobile app carries the same view.)

The overlay title reads Triage Log under the Decisions eyebrow. Below it:

Decisions the system auto-archived to keep the feed manageable. Click Restore to bring one back.

The summary line

Right under the description, four counts run left to right:

• Last 7 days: how many decisions the system has archived in the past week.
• Last 30 days: how many it has archived in the past month.
• By rule engine: how many of those 30 days' archives came from the deterministic nightly rules.
• By agent: how many came from the weekly agent pass.

One quirk worth knowing: all four of these counts look back over the same 30-day window, so By rule engine plus By agent adds up to Last 30 days, and Last 7 days is a subset of that. The row list underneath goes back further, 60 days, so it's normal to scroll down and see rows older than the 30-day window the counts summarize. The counts tell you the recent pace; the list is the full 60-day record.

Reading a row

Every archived decision is one row, separated from the next by a hairline, no card around it. Each row shows:

• A colored dot for severity: gray for Info, yellow for Warning, red for Critical. Same severity scale used everywhere else on Decisions.
• The decision's title, exactly as it read before it was archived.
• An archive-reason label (see the full list below), telling you which rule or which agent judgment archived it.
• The section it came from, if known, humanized (Vendor, Carrier, Margin, Job, and so on), followed by the specific entity's name if the decision was tied to one, a vendor, a carrier, a client.
• How long ago it was archived, in relative time: hours, days, weeks, or months.
• "Restored before", in amber text, if you have already restored this exact decision at least once and the system re-archived it again since. This is Verinode quietly flagging that this particular rule may be miscalling this particular case repeatedly, it's tracked to help tune the rules over time, not just decoration.
• A quoted note in italics, only on agent-archived rows, when the agent left a short written reason for its call, something like why it judged a plan stalled or a decision low priority that week.
• A Restore button, a small pill on the right of every row.

Archive reasons

Every row is archived for exactly one of these seven reasons. The first five come from the nightly rule engine, the last two from the weekly agent pass:

| Label you see | What it means | |---|---| | Low-impact, idle 30+ days | An Info-severity decision that has sat untouched for 30 days or more. | | Mid-impact, idle 60+ days | A Warning-severity decision idle for 60 days or more. | | Parked, idle 90+ days | A decision you had explicitly parked that then sat for 90 days or more with no further action. | | Past expiration | The decision's own window for acting on it has passed. | | Duplicate of newer signal | A newer decision superseded this one; Verinode tracks internally which one replaced it. (Verinode's internal name for a decision record is a "signal," which is why this particular label uses that word instead of "decision.") | | Agent: low priority this week | The weekly agent pass judged this decision not worth your attention this particular week. | | Agent: stalled plan | The weekly agent pass judged the action plan behind this decision had stalled, no real progress since it started. |

Critical-severity decisions are never swept up by the age-based rules. The only way a Critical decision gets archived is through the agent's own judgment, and even then it has to be genuinely stale first.

Restoring a decision

1. 1Open Decisions from the sidebar, then click Triage Log in the top right.
2. 2Scan the list for the decision you want back. Sort by eye, most recent archives sit at the top.
3. 3Click Restore on that row.
4. 4The button reads Restoring… for a moment while the request goes through.
5. 5On success, the row disappears from the triage log immediately, no page reload needed. The decision is back on Decisions and the Feed within moments.

Restoring flips the decision's status back to active and clears its archive record, which is exactly why it drops out of this log the moment you restore it: the log only ever shows decisions carrying an active archive record. Verinode also keeps a quiet history of the restore, which rule or agent call it reversed and when, so a decision that keeps getting archived and restored builds a track record that feeds back into how the rules get tuned.

If a restore fails, you'll see a short red message under that row instead, explaining why (for example, if the decision was already dismissed some other way in the meantime). It's rare: because the log only lists decisions the system itself archived, almost every row here is restorable.

Restore works one decision at a time. There's no bulk "undo everything this rule caught" button on this screen today, if several rows share a reason you'd like reversed, restore them individually.

Empty state

If nothing has been auto-archived in the last 60 days, the log reads:

Nothing has been auto-archived recently. Triage runs nightly. Check back tomorrow.

That's a quiet system, not a broken one. It means your open decisions have all stayed current enough that neither the nightly rules nor the weekly agent pass has needed to clear anything out.

Best-practice example

Say the summary line reads Last 7 days: 3, Last 30 days: 12, By rule engine: 9, By agent: 3. Scanning down, most rows read Low-impact, idle 30+ days, ordinary housekeeping on Info-severity decisions nobody acted on. One row stands out: Mid-impact, idle 60+ days, tied to a carrier you're still actively negotiating with, with Restored before already showing in amber. That combination, a decision tied to something live, restored once already, is worth a second look: click Restore, and it's back on your Decisions list in seconds, ready to act on again.

Related reading

• The decision workspace: what opens when you act on any decision, restored or not.
• Acting on decisions: what Pending, Parked, Acted, Ignored, and Resolved mean.
• The Feed: where restored decisions reappear alongside everything still active.