Usage and Add-ons

Understand Cloud Agents quotas, usage analytics, throughput signals, and add-ons that expand review, execute, custom-agent, and parallel capacity.

Cloud Agents has two capacity views that belong together:

  • Usage tells you what the workspace is consuming now.
  • Add-ons lets you buy more room when demand outgrows the current plan.

Who Should Read This Page

RoleWhat they usually need from Usage
Engineering managersWhether review and execute capacity matches team demand
Platform teamsWhether friction comes from quota, policy, or configuration
Finance or operations stakeholdersWhether Cloud Agents demand is steady, bursty, or tied to specific release periods
Repository ownersWhether one repository is using more capacity than expected

Usage

The Usage tab turns Cloud Agents from a black box into an operating surface with measurable throughput.

Filters

The page includes filters such as:

  • date range
  • status
  • run type
  • repository

These filters let teams answer whether the issue is recent, specific to execute runs, or concentrated in one repository.

Core quota cards

CardWhat it means
Review QuotaReview runs consumed in the current billing period versus the current limit
Execute QuotaExecute runs consumed in the current billing period versus the current limit
Daily Run QuotaTotal Cloud Agent runs created today against the daily cap

The daily cap matters operationally because it prevents a burst of activity from overwhelming the workspace in one day.

Operating metrics

Additional summary cards usually show:

MetricWhat it tells you
Runs in rangeTotal run volume for the selected period
Success rateShare of runs that completed cleanly
Total tokensToken consumption for the selected slice
Avg durationTypical runtime and whether work is slowing down

What these metrics mean beyond the UI

TrendUsually indicates
Review quota rises faster than execute quotaThe organization is using Cloud Agents mainly as a review accelerator
Execute quota rises with stable success rateThe team is trusting controlled implementation workflows more often
Token usage grows faster than run countPrompts, context size, or model selection may be getting heavier
Average duration spikesTasks are becoming broader, repositories are more complex, or the selected models are mismatched to the job

Charts and breakdowns

Daily Activity

Shows how run volume moves over time and splits the activity by review, execute, and investigative work where available.

Status Breakdown

Shows how much work is completing, failing, awaiting approval, queued, or cancelled.

Top Repositories

Shows where Cloud Agents activity is concentrated so teams can identify hotspots or uneven adoption.

Trigger Sources

Helps separate manually created work from event-driven or system-driven runs.

Reading Usage Operationally

Healthy pattern

  • review quota rises steadily
  • success rate remains high
  • awaiting-approval does not dominate the status chart
  • top-repository concentration matches team expectations

Warning pattern

  • daily run quota is maxed early in the day
  • queued or awaiting states dominate
  • failure rate rises after a repository-policy change
  • one repository consumes most workspace capacity unexpectedly

How to Use Usage in Operations Reviews

Weekly

  • check whether review and execute volume matches actual team priorities
  • identify repositories that are over-consuming capacity or producing noise
  • compare success rate changes against recent repository-policy edits

Monthly

  • decide whether recurring add-ons are justified
  • determine whether daily caps are consistently constraining the team
  • compare usage concentration against repository importance and staffing

Add-ons

The Add-ons tab expands workspace capacity. The exact catalog can vary, but the current product surface includes three main shapes:

Add-on typeWhat it does
RecurringAdds monthly capacity to the workspace on an ongoing basis
One-timeAdds a single top-up for short bursts or temporary needs
MeteredCharges based on actual incremental usage instead of a fixed bundle

Common add-on families

The workspace catalog can include items such as:

  • Review capacity extensions for more review runs
  • Execute capacity extensions for more code-change runs
  • Profile capacity extensions for more custom agent slots
  • Flex / PAYG review overage options
  • Parallel collaboration token packs for higher multi-agent concurrency

The Console is the source of truth for current pricing and packaging.

Choosing the Right Add-on

SituationBest response
Your team consistently hits review quota every monthChoose a recurring review extension
A release period temporarily increases demandChoose a one-time top-up
You only rarely exceed quota but want overflow protectionChoose a metered option where available
More teams need specialized agentsAdd custom-agent profile capacity
You want more concurrent collaborationAdd parallel collaboration tokens

When Not to Buy More Capacity Yet

Do not reach for add-ons first when the real problem is quality or governance.

  • If failure rates are climbing, fix prompts, models, or repository policy first.
  • If awaiting approval dominates, the bottleneck is human workflow design, not quota.
  • If one repository is generating most of the load, tune that repository before buying workspace-wide capacity.
  • If token usage is rising without clear value, improve task scope before paying for more throughput.

Budgeting and Team Planning

Usage and add-ons matter beyond engineering throughput:

  • platform teams use them to keep automation predictable
  • engineering managers use them to plan review coverage
  • finance teams use them to understand whether automation demand is steady or bursty

The right question is not only "how much are we using?" It is also "are we using Cloud Agents on the work where it produces the most leverage?"

For broader account-level plan and billing context, see Account Plans and Billing.

How is this guide?

On this page