Skip to content
Activity log

Know what happened before the client asks.

Every deploy, DNS edit, automation run and invitation lands in one feed as it happens, signed by whoever did it, human or robot. One filter turns it into the answer.

All projects Atelier Nord Loopline + 10 projects
06:41 Ubuntu updates finished on 8 servers
09:12 NR Deployed loopline.app
10:05 Costs synced from OVHcloud
11:26 ML Updated an MX record on atelier-nord.fr
14:12 ML Pushed a1f3c92 · build passed
15:03 Issue found: DMARC missing on galerie-ondine.fr
16:30 LF Invited Kenji Tanaka to the team
17:15 NR Updated the deploy script on atelier-nord.fr
It wrote itself. All day.
What is the Agency Activity Log in Unolia?

The Agency Activity Log is one chronological record of everything that happens across your client projects: 41 kinds of events covering deployments, DNS edits, automation runs, provider syncs, monitor changes, team and role changes, commits, issues found and fixes applied. Each entry is signed by its actor, a teammate, an automation or a provider, and the feed filters by project, by kind of event and by source.

The record

Everything that happens, happens on the record.

An agency day is written by many hands: teammates deploying and editing, automations doing the rounds, providers syncing in the background, Unolia flagging what it finds. All of it lands in the same feed, in order, as it happens.

  • Deploys, edits and invitations, signed by teammates.
  • Automation runs and fixes, filed by the robots.
  • Provider syncs and imports, logged as they land.
What gets recorded 41 kinds of events
Deployments DNS Automations Providers Monitors Team and roles Repositories Issues and fixes Costs
Signed by a person, a robot or a provider
How it works

From noise to a lookup

1

Let it write

Every action lands in the feed as it happens, whoever or whatever performed it. Nobody files a thing.

2

Filter to the question

One project, one kind of event, one glance. The client question becomes two clicks.

3

Answer with the record

Support calls, renewal meetings and post-mortems all read from the same page.

The robots

The robots keep their own diary.

Automations, auto-fixes and even their reverts land in the same feed as human work, timestamped and named. What did the machine do last night is a question the machine already answered.

Written by the robots this week
Ubuntu updates finished on 8 servers Tue · 06:41 · automation ✓
Auto-fix: www CNAME recreated on maisonverdier.com Wed · 04:12 · auto-fix ✓
Auto-fix reverted: SPF change rolled back on shop.brand.fr Wed · 04:15 · auto-fix ↶
Even the reverts are on the record
DNS · Atelier Nord filtered
ML MX record updated Marc Lefebvre · today · 11:26
NR SPF record added Nadia Roche · Mar 3
Zone imported from Gandi Provider · Feb 12
Asked and answered in one filter
The question

Who changed this? stops being an investigation.

Pick the project, pick the kind of event, and the answer is a short list with names and times. No channel archaeology, no asking around, no wondering whether it was a person or a sync.

In and out

Plugged in at the source. Queryable at the other end.

Unolia listens to your providers over webhooks, so events land in the record the moment they happen, not the next time someone syncs. And when you need the history back out, take it your way: through the API, or by asking your AI agent in plain language.

The record · in and out live
In · webhooks
Push received from GitHub instant
Deploy finished on Forge instant
Monitor status change from Oh Dear instant
Out · on demand
API GET /activity?project=atelier-nord
AI What changed on Atelier Nord this week?
In the moment it happens. Out the moment you ask
In real life

The moments this changes

The support call

The client says the site behaved oddly on Tuesday. You filter to their project while they are still talking, read the 11:26 DNS edit out loud, and the mystery is over before the call is.

The morning after

You were off yesterday. The feed reads like a diary of everything the team and the automations did while you were gone. Caught up before the first coffee.

The blameless post-mortem

Something broke at 03:12. The record shows what ran, what changed and in which order, in minutes and names instead of memories. The fix follows facts.

How this grows with your plan

See pricing →
Developer

Every event recorded from day one.

Freelancer

The per-project activity feed.

Startup

One team-wide feed across every project.

Business

50 seats writing to the same record, custom roles around it.

Related feature pages

Team Access Match every provider account to a person, scope access with roles and groups, catch drift. Automations Ready-made maintenance recipes that run across every server, pause for your yes and keep the whole story. Repository Insights Commits, build runs and dependency changes, turned into a story anyone at the agency can read.

Stop searching Slack to know what happened.

Join the waitlist and give the whole agency one record to read.

Join the waitlist