NEW Scheduled Agents — the recurring worker

AI you ask becomes AI that brings you things.

A Scheduled Agent wakes on your cadence — every weekday at 8, Mondays at 9, the first of the month — grounds itself in your workspace memory, does a defined job, and leaves a durable result. Every single time.

A recurring workspace worker — not a reminder. It has a persona, a home channel, an output, a budget, and an auditable trail.
cron · 0 8 * * 1-5America/New_York · weekdays 08:00Medium tierad-report
01 · WAKE

The alarm fires

On its cron schedule, the agent wakes — reserving its run budget before it does anything.

queued → running
reserved · Medium tier
02 · GROUND

Loads your memory

It grounds in the workspace — profiles, indexes, the target artifact, and recent run history.

profile + #growth
last 7 versions
03 · DO THE JOB

Runs its task

It pulls live data, reasons, and produces the output it was configured for — under its caps.

meta_ads · stripe
9 tool calls · $0.18
04 · LEAVE A RESULT

Durable output

A new artifact version or a thread — announced in your channel, with a full trace.

✓ v8 · daily-ad-report
✓ posted to #growth
The mental model

Four ways an agent wakes up.

Boardbox agents share the same memory, governance, and budgets. What differs is the trigger. Scheduled is the recurring worker — the one that runs on a clock, whether you’re watching or not.

live

Channel

Wakes when you send a message.
Like a teammate in a channel.
trigger · you, in conversation
this

Scheduled

Wakes when a cron timer fires.
An alarm clock. A standing job.
trigger · a cadence you set
live

Reactive

Wakes when an external event arrives.
A pager. A nervous system.
trigger · a webhook, in real time
live

Deep

Wakes when you hand it a big task.
A contractor you brief once.
trigger · an objective

Scheduled and Reactive are close cousins — both autonomous, both governed, both leave a run record. The difference is what wakes them: Scheduled runs on an alarm clock, asking “anything change since yesterday?” Reactive runs on a nervous system, told “this just changed” the instant it does. Scheduled is your operating cadence; Reactive is your reflexes.

Two output modes

Bring me a document, or start the conversation.

Every Scheduled Agent does one of two things with its run. The mode is the difference between a living artifact and a recurring thread.

Report mode

It brings you a document.

The agent produces or updates a living artifact. Each run is a new version— so the artifact’s own history becomes the agent’s long-term memory of every prior run.

daily-ad-report.mdv8 · today
Daily ad performance — Jun 2
Spend (24h)$138
Blended ROAS0.7x
Best campaignretargeting 1.9x
historyv5v6v7v8
Conversation mode

It starts the conversation.

The agent posts into a channel and opens a thread with a visible headline. Best when your reply matters — check-ins, journaling, standups, retros, recurring questions.

#health · morning check-in
Rx
Rex08:00
Morning, Clayton. You logged 6h 10m of sleep — third short night this week. How’s your energy for training today?
Reply to continue the thread…
Why it’s different

It wakes up inside a workspace, not a blank prompt.

A normal scheduled reminder fires a static message. A Scheduled Agent grounds every run in durable memory — so it knows what changed since last time.

Grounded in your memory.

Before it does the job, the agent loads what it needs: your personal profile, the workspace and channel index, the target artifact and its recent versions, pending proposals, and — for governed monitors — the plan’s watch conditions.

Continuity across runs.The artifact’s version history is the agent’s memory of what it found before.
Not a reminder. It reasons over live context, then produces something new each time.
run #142 · grounding
Personal profileloaded
Channel index · #growthloaded
Target · daily-ad-report.mdv7 read
Recent versions · last 7diffed
Live read · meta_ads200 OK
Four ways in

However the idea arrives, it becomes a job.

You always review the proposed setup — name, schedule, mode, persona, tools, channel, and budget — before the agent exists.

1

Blank brief

Describe what you want, when it runs, and what it should look at. Boardbox generates a starting configuration you can tune.

“Every weekday at 8, summarize ad spend and flag anything bleeding.”
2

Suggested for you

Boardbox reads your Personal profile or team Workspace Brief and proposes tailored agents — each tied back to context it already knows.

Suggested · “Weekly investor update, Mondays 9am” — from your Workspace Brief.
3

Make a thread recurring

Turn an already-useful conversation into a scheduled job. The thread summary becomes the seed brief.

“Make this standup recurring” → every Monday, same crew, same questions.
4

Governed monitor from a plan

A planning flow proposes a monitor tied to a plan — what to watch and when to escalate. The most autonomous path.

Watch the Q3 plan’s kill criteria · escalate if payback > 6mo.
Bounded by design

Every run has a budget. Every run leaves a trace.

You pick a tier, not a token count. Runs reserve budget before executing and settle to actual cost after — and if quota’s short, the run is skipped and surfaced, never failed in silence.

Light
Quick check-ins and prompts.
cost~$0.05
tools≤ 3
time30s
Medium
Daily reports and briefs.
cost~$0.25
tools≤ 12
time2m
Heavy
Deep weekly research.
cost~$1.50
tools≤ 40
time8m
Max
Monthly strategy snapshots.
cost~$5.00
tools≤ 100
time20m
queuedrunningsucceededskipped · quotafailed 3 fails → auto-pause
Governed monitors

It shows its work before it earns autonomy.

The deepest agents run under an explicit autonomy policy: what auto-applies, what must escalate, daily action ceilings, spend caps, and an expiry. It doesn’t ask you to trust it — it shows what it would do, and how recent history would have behaved under the policy.

Reversible work runs free. Versioned writes inside Boardbox; destructive or external actions gate.
Re-engages on breach. Exceed the envelope and it returns to you with a structured diagnosis and an approval path.
Autonomy policy · Meta Dailymonitor · what runs vs what gatesactive
Refresh ad copy on fatigued setsauto
Rotate to approved creativeauto
Reallocate spend > $50/daygate
ceiling 8/day · cap $200/mo · expires Sep 30
Not a reminder on a timer

AI on a timer, inside an operating system.

The moat isn’t “AI runs on a schedule.” It’s that the schedule fires inside a workspace with durable memory, channel placement, real output, and approval boundaries.

Proactive chat / tasks
Closer to a scheduled reminder — fires a prompt, gives you a message.
No durable memory of prior runs. Each fire starts fresh.
No artifact, no channel home, no version history.
No budget tiers, run records, or approval boundaries.
Boardbox Scheduled Agent
A recurring job with a persona, schedule, home channel, and output mode.
Grounded in memory — it knows what changed since last run.
Produces a versioned artifact or a structured thread, every time.
Budgets, full run history, and governed autonomy with approvals.
Operating cadences

The standing jobs that run your week.

weekdays · 08:00
Daily ad performance report
every morning
Pulls overnight spend and ROAS, updates the living report, and flags anything bleeding in #growth.
report modeMedium
mondays · 09:00
Weekly investor update
every week
Compiles metrics, wins, and asks into a memo — versioned weekly so the whole quarter is one history.
report modeHeavy
daily · 08:00
Morning health check-in
every morning
Reads last night’s sleep, opens a thread with a tailored question, and keeps the streak going when you reply.
conversation modeLight

Put your week on autopilot.

Describe a Scheduled Agent in plain language, pick its cadence, and let it bring you what matters — grounded, governed, and on time.