Flags
An Agent Serve flag is a true-or-false rule that decides which rows appear in an analysis.
Renewal in 30 days AND usage declined
Agent Serve · Rules
Business logic everyone can follow.
Define “at risk” or “qualified” once. Every dashboard, table and agent checks records the same way, and shows why a row passed.
Agent Serve Rules makes business definitions explicit in Agent Serve, the AI workspace where teams build, deploy, and manage AI agents. Rules are reusable flags, classifications and scores built from conditions and groups, used to filter records, derive columns, supply states for dashboard values, drive monitors and guide Relays.
How it works
Write it once, use it everywhere.
An Agent Serve rule is a name, a type and a condition. Once saved, dashboards, monitors and Relays all use the same definition.
- 1
Name the rule
Give it a name people search for, such as High Risk Opportunity, and an optional description.
- 2
Pick the type
Choose Flag (true / false), Classification (labels) or Score (points).
- 3
Build the condition
Combine conditions in all or any groups, with operators such as between, in, contains and is empty.
- 4
Use it everywhere
Filter dashboard tables, colour maps, set KPI states, watch matches in Monitors, and bind it to a Relay.
Three kinds
Flag it, class it, or score it.
Classifications
An Agent Serve classification sorts records into labelled buckets, each with its own condition, plus a fallback label.
Healthy · Watch · At risk
Scores
An Agent Serve score adds points when each condition holds, so the team can read and edit how a number is built.
+3 points when overdue > 30 days
Where rules run
One rule, six places it counts.
Agent Serve uses the same rule in dashboard tables, KPI states, maps, monitors, Relays and call compliance.
01 · Dashboard tables
Include, exclude, explain
An Agent Serve dashboard table can include or exclude rows with a rule, or add a rule result column beside the data.
- Keep only flagged opportunities in a review table
- Show the class or score as its own column
- Change the rule and every table follows
02 · KPI states and maps
Colour that means something
Agent Serve rules set the state badge on a KPI card and colour map regions, so colour follows a definition rather than taste.
- Outcome labels can carry their own colours
- The same rule on a card and on a map
- Readers see the state, not just the number
03 · Call compliance
A breach means one thing
In Agent Serve call compliance, a breach rule decides which failed checks count as a breach in the report.
- Each check marked with its severity
- The breach rule applies to every call
- Reviewers see the reason in plain language
04 · Monitors and Relays
Alerts and agents that agree
An Agent Serve Rule match monitor alerts when enough rows match a rule, and a Relay bound to that rule checks records the same way.
- Rule match counts rows in a workspace table
- Relays take rules as resources
- One definition from alert to follow-up
Examples
Rules banks and insurers write down once.
Three definitions financial services teams keep in Agent Serve Rules instead of in spreadsheets and people’s heads.
High-risk opportunity flag
An Agent Serve flag marks large deals in late stages with a weak credit rating, and the pipeline review table shows only those.
- Flag rule
- Dashboard table
- Rows included
- Credit review
AML threshold flags
An Agent Serve flag marks transactions at or above the reporting threshold, and a Rule match monitor alerts when enough rows match.
- Transactions
- Threshold flag
- Rule match
- Analyst alert
Complaint classification
An Agent Serve classification sorts complaints into Regulatory, Service and Billing, with General as the fallback label.
- Complaints table
- Classification
- Labelled buckets
- Dashboard by class
Examples show what teams can build in Agent Serve. They describe patterns, not named customer deployments.
Reuse
One definition, every view.
A rule lives apart from the panels that use it. Change the definition once and every dashboard, table and monitor follows.
Under the hood
Agent Serve Rules, in detail.
| Rule types | Flag (true / false), Classification (labels) and Score (points). |
|---|---|
| Groups | Conditions combine in all, any and not groups. |
| Operators | =, ≠, >, ≥, <, ≤, in, not in, between, contains, is empty and is set. |
| Classifications | Labelled buckets, each with its own condition, plus an optional fallback label. |
| Scores | Scoring terms that add points when their condition holds. |
| Outcome colours | Outcome labels can carry colours for badges and maps. |
| Used by | Dashboard tables, KPI state badges, map colouring, Monitors (Rule match), Relay bindings and the call compliance breach rule. |
Before and after
Definitions people stop arguing about.
| Work | Before | With Agent Serve |
|---|---|---|
| What “at risk” means | BeforeA formula copied into every report, each slightly different. | With Agent ServeOne Agent Serve rule, used by every dashboard, monitor and Relay. |
| Changing the definition | BeforeHunt down every report and update it by hand. | With Agent ServeEdit the rule once, and every place that uses it follows. |
| Explaining a score | BeforeA number nobody can trace back to its inputs. | With Agent ServeScoring terms anyone can read: points, and when they apply. |
By the numbers
Small rules, wide reach.
- Rule types: flag, classification and score
3
Rule types: flag, classification and score
- Operators, from equals to is set
12
Operators, from equals to is set
- Group kinds: all, any and not
3
Group kinds: all, any and not
- Places a rule is used across Agent Serve
6
Places a rule is used across Agent Serve
Works with
Rules in context.
Further reading
Solutions built on Agent Serve Rules.
- Solution deep dive
Reconcile failed card disputes before the chargeback deadline
Disputes that fail reconciliation tend to sit in a side queue until the chargeback window has closed. Here is an end-to-end setup that finds them early, ranks them by deadline and drafts the pack, with an analyst signing every filing.
Cards and payments operations · 12 min read - Solution deep dive
Clear the KYC periodic review backlog without adding analysts
Periodic reviews pile up because every file starts from scratch, even when nothing has changed. Here is a setup that refreshes the data, drafts the file and routes it by risk, so analysts spend their time on the reviews that need judgement.
Retail and business banking · 12 min read - Solution deep dive
Catch missing risk warnings on recorded advice calls within 24 hours
Sampled call reviews find a missing risk warning weeks after the client acted on the advice. Here is a setup that checks every recorded advice call within a day and puts the quote in front of a compliance reviewer.
Wealth and advice · 12 min read
Questions
About Rules
How do rules affect dashboards?
Agent Serve rules can include or exclude records, derive columns, classify map features and supply states for metric displays. A rule’s definition lives separately from the panels that use it.
What is the difference between a flag, a classification and a score?
In Agent Serve, a flag is true or false, a classification sorts records into labelled buckets with a fallback label, and a score adds points when each condition holds.
Can a rule trigger an alert?
Yes. An Agent Serve Rule match monitor fires when at least the number of rows you set in a table match the rule.
Can AI agents use rules?
Yes. An Agent Serve Relay can take rules as resources, so it checks records with the same definitions as your dashboards and monitors.
Last updated