Skip to main content
Updated Mar 13, 2026

RevOps Agents and the Revenue Dashboard

In Lesson 12, you built jurisdiction-compliant outreach for three markets. Every workflow in this chapter so far has required you to type a prompt. Research, score, enrich, draft outreach, analyse campaigns, check compliance -- all on demand, all dependent on you remembering to run the right command at the right time.

That works when you have fifteen prospects and one rep. It breaks when NexaFlow has fifty active deals, three markets, and a sales team that needs pipeline intelligence before their first coffee. What if the system ran these workflows automatically, on schedule, without waiting for anyone to remember?

RevOps agents run on cadence. They monitor signals, enrich CRM data, manage sequences, analyse campaign performance, and report on pipeline health. They produce digests. They never take autonomous action. The rep who sees a fresh signal, reviews a revenue report, and decides to call Meridian Logistics first is exercising judgment the agent cannot replicate. The agent surfaced the change. The human chose the response.

In the actual sales-revops-marketing plugin, you work with two asset types:

  1. Skills in skills/, such as pipeline, crm-enrichment, sequence, and pre-call-brief
  2. Agents in agents/, such as lead-intelligence-agent and revenue-reporting-agent

Cowork can invoke either one from a natural-language prompt. The practical difference is scope: skills handle bounded tasks, while agents represent an operating role with a cadence, checklist, and output format.

Real-Life Constraint

In this chapter, you already have the working data context: Lesson 1 generated demo-data.md, and the folder instructions tell Cowork to keep using it throughout the chapter. To make these agents operational in Cowork, you mainly need two things: sales-marketing.local.md for ICP and pipeline rules, and /schedule when you want Cowork to run an agent task later on a cadence instead of immediately.

The Five RevOps Agents

Each agent serves a distinct operational function. Together, they form a coordinated monitoring and reporting system that keeps NexaFlow's revenue engine running without manual prompts.

How To Read This Lesson

For each agent below, do two things:

  1. Prompt it directly in Cowork so you see what the asset actually does
  2. Decide where it belongs in the weekly operating rhythm

This keeps the lesson grounded in the real plugin: named skills, named agents, and prompts that invoke those actual assets.

Agent 1: Lead Intelligence Agent

The Lead Intelligence Agent watches for buying signals across your pipeline and alerts reps when something changes.

PropertyConfiguration
MonitorsFunding announcements, leadership changes, contract wins, hiring surges, social media posts
Related skillsprospect-research + lead-scoring
TriggerAny HOT signal generates an alert within 2 hours
ScheduleDaily scan at 06:00 UTC, weekly digest every Monday
OutputPriority-sorted signal digest with score impact
Prompt it directlyUse the lead-intelligence-agent to scan NexaFlow's watch list for new HOT or WARM signals since yesterday. Update scores and recommend the next rep action for each HOT signal.

When DataForge Solutions in Karachi posts two engineering roles mentioning "logistics integration" on Rozee.pk, the agent flags it as HOT. When Gulf Express in Dubai's VP of Operations posts about modernising fleet management on LinkedIn, the agent catches it within hours. Neither signal required a rep to remember to check.

Agent 2: CRM Hygiene Agent

The CRM Hygiene Agent scans your database for data quality issues that degrade scoring accuracy and pipeline reliability.

PropertyConfiguration
MonitorsStale records (no update in 30+ days), duplicate entries, conflicting data across sources
Related skillscrm-enrichment + lead-scoring
ActionAutomatic enrichment of outdated fields, role change flagging, duplicate detection
ScheduleWeekly for Tier 1 accounts, monthly for all accounts
OutputCategorised issue report with recommended actions
Prompt it directlyUse the crm-hygiene-agent to audit NexaFlow's HOT and WARM leads with last enrichment older than 7 days. Refresh stale records, flag role changes, and report the data-quality issues that need rep review.

A CRM where Kaizen Supply Co. shows 120 employees while LinkedIn shows 340 produces inaccurate Fit scores. The Hygiene Agent catches the discrepancy and recommends the update. A rep who relies on the old figure underestimates the deal size and under-invests in the relationship.

Agent 3: Outreach Sequencing Agent

The Outreach Sequencing Agent monitors active sequences and manages touch progression based on prospect behaviour.

PropertyConfiguration
MonitorsSequence progress, email opens, link clicks, replies, bounces
Related skillssequence + outreach
ActionTriggers next touch on schedule, pauses on reply, stops on bounce or opt-out
ScheduleContinuous (event-driven, not time-driven)
OutputSequence status report with engagement metrics
Prompt it directlyUse the outreach-sequencing-agent to review NexaFlow's active HOT and WARM sequences. Show which prospects need the next touch, which sequences must pause because of replies or bounces, and what the next approved touch should be.

This agent is different from the others. It runs continuously, reacting to events rather than waiting for a schedule. When a prospect opens email three in a six-touch sequence and clicks the case study link, the agent advances to touch four. When a prospect replies "not interested," the agent pauses the sequence immediately. When an email bounces, the sequence stops and the CRM record gets flagged for the Hygiene Agent to investigate.

The over-automation risk from Lesson 6 applies here. An agent that sends touch five to a prospect who replied "let me think about it" after touch three has ignored a human signal. Exit conditions matter: reply (any reply) pauses. Bounce stops. Opt-out stops permanently. The agent handles timing and progression. You handle the judgment calls.

Agent 4: Marketing Performance Agent

The Marketing Performance Agent evaluates campaign results across channels and identifies where budget is working and where it is not.

PropertyConfiguration
MonitorsLinkedIn Ads, email nurture, content/SEO, Google Ads, events
Related skillsperformance-analysis
ActionWeekly analysis report with channel comparison, CPL trends, and optimisation recommendations
ScheduleEvery Friday
OutputChannel performance digest with budget status
Prompt it directlyUse the marketing-performance-agent to analyse NexaFlow's last 7 days of campaign performance against targets. Compare to the prior week and recommend the top 3 optimisations.

NexaFlow's email nurture campaign delivers leads at $9.50 CPL with 4.8% conversion. Google Ads delivers leads at $387.50 CPL with 0.9% conversion. The Marketing Performance Agent flags the disparity. But the recommendation to "pause Google Ads" requires context the agent does not have -- if Google Ads targets enterprise logistics directors while email targets mid-market operations managers, comparing CPL across them is misleading. The agent compares numbers. You compare strategy.

Agent 5: Revenue Reporting Agent

The Revenue Reporting Agent aggregates pipeline metrics, tracks deal velocity, and produces the weekly revenue dashboard.

PropertyConfiguration
MonitorsPipeline value, conversion rates, deal velocity, forecast accuracy, stage stagnation
Related skillspipeline
ActionWeekly revenue dashboard with executive summary
ScheduleEvery Monday
OutputDashboard with seven core metrics plus five-bullet executive email
Prompt it directlyUse the revenue-reporting-agent to build NexaFlow's weekly revenue dashboard. Include pipeline coverage, at-risk deals, lead velocity, marketing contribution, and forecast versus quarterly target.
Connector Note

For this chapter, you do not need Slack, email, or other external delivery tools to make the lesson work. If you run /schedule in Cowork and assign one of these agent tasks, Cowork executes that task at the scheduled time using the workspace context you already set up in Lesson 1. External CRM or calendar connectors can enrich the workflow later, but they are not required for the core pattern here.

The pipeline Skill: Analysis and Forecasting

NexaFlow's pipeline has ten active deals. The main skill asset for deal analysis in this lesson is pipeline. You can invoke it directly from Cowork, or, if the base Sales plugin is installed, use /pipeline-review and /forecast as familiar entry points into the same working context.

Prompting the pipeline Skill

Use the `pipeline` skill to review NexaFlow's active pipeline.
Show deal-level health scores, weighted pipeline, stalled deals,
and realistic forecast this quarter.

If the base Sales plugin is also installed, Cowork may surface similar analysis through /pipeline-review and /forecast. The important plugin asset in this lesson is the extension's pipeline skill, which adds three-dimension scoring integration (Fit + Timing + Engagement from Lesson 3) and deal-level health logic.

What to expect: The pipeline skill produces a deal-level health assessment. Your output will vary, but look for these sections:

SectionIntentWhat to Verify
Deal health scoresPer-deal score with Fit/Timing/Engagement breakdownEach deal scored on three dimensions from Lesson 3
Stage and riskCurrent pipeline stage with risk rating (LOW/MEDIUM/HIGH)High-risk deals show long stage duration or low engagement
At-risk dealsDeals requiring immediate attentionStalled deals, silent contacts, missing champions flagged
Strongest closesDeals most likely to close this quarterHigh health scores with active engagement
Forecast gapWeighted pipeline vs quarterly targetShows whether current pipeline covers the target
Your output will vary

The pipeline data depends on your demo-data.md content and the deals you have been working throughout this chapter. The teaching point is interpreting deal health in context: a deal stalled in Proposal for 30+ days with no engagement is not "in progress" -- it is silent. A deal with high Fit but low Timing may close next quarter, not this one. The agent scores deals on data it can measure. You evaluate deals on context it cannot.

Review the output for three categories: deals that need immediate intervention (stalled, silent, or missing a champion), deals that are your strongest closes (high health, active engagement), and the gap between your weighted pipeline and quarterly target.

Weighted pipeline calculation must show the per-deal math: multiply each deal value by its stage probability (Discovery=20%, Qualification=40%, Proposal=60%, Negotiation=80%), then sum all weighted values. For each stalled deal, calculate what percentage of the remaining quota gap that deal represents -- this quantifies the impact of losing it and determines whether replacement pipeline is needed.

Meridian Logistics and Gulf Express are the strongest closes. Meridian has a champion (Sarah Chen), an active proposal, and scores of 92/100. Gulf Express is in negotiation with high engagement. These two deals represent $275,000 in near-term revenue.

Sales Forecast

With the pipeline reviewed, use the same pipeline skill to model revenue scenarios.

Forecast Prompt

Use the `pipeline` skill to forecast NexaFlow's current quarter.
Show best, likely, and worst case scenarios.

If the base Sales plugin is installed, compare this with `/forecast`.

/forecast belongs to the base Sales plugin. In the chapter stack, it complements the extension's pipeline skill. In direct plugin use, you can still ask the pipeline skill for committed, likely, upside, or best/likely/worst scenario logic.

What to expect: The pipeline skill or /forecast produces scenario-based revenue projections. Your output will vary, but look for these sections:

SectionIntentWhat to Verify
Per-deal probabilityBest/likely/worst case for each dealProbability reflects deal stage, engagement, and signal strength
Scenario totalsAggregate revenue for each scenarioBest case shows ceiling, worst case shows floor
Target comparisonScenarios vs quarterly targetShows which scenario meets target and which falls short
Your output will vary

The forecast depends on your pipeline data. The teaching point is evaluating assumptions: Does the agent's probability estimate for your largest deal match your knowledge of the procurement cycle? Is a 40% "likely" probability realistic for a deal with a slow decision-maker? The gap between your assessment and the agent's is where your judgment adds value. The agent calculates. You contextualise.

Evaluate the assumptions. For your swing deals -- the ones where probability estimates determine whether you hit target -- ask whether the agent's probability reflects what you know about the procurement timeline, competitive dynamics, and champion strength. That judgment is yours, not the agent's.

Building the Revenue Dashboard

The revenue dashboard consolidates pipeline health, forecast, and operational metrics into a single weekly view. In Cowork, you can invoke the revenue-reporting-agent directly with a prompt. That agent then frames the reporting role for Monday morning RevOps.

#MetricSourceWhat It Reveals
1HOT leads generated (this week)Lead Intelligence AgentInbound signal quality
2Lead-to-SAL conversion rateCRM + Scoring dataQualification effectiveness
3Pipeline created (this week)CRM new opportunitiesGrowth trajectory
4Average deal sizePipeline dataMarket positioning
5Pipeline at risk ($ value)pipeline skill / /pipeline-reviewRevenue exposure
6Close rate (trailing 90 days)CRM closed-won/closed-lostSales effectiveness
7CAC by channelMarketing Performance AgentBudget efficiency

Running the Dashboard

Use the `revenue-reporting-agent` to build NexaFlow's weekly revenue dashboard.
Include all seven metrics below. Then produce a weekly executive email
in five bullets, maximum 150 words.

What to expect: The agent produces a weekly dashboard with metrics and executive summary. Your output will vary, but look for these sections:

SectionIntentWhat to Verify
Seven metrics tableThis week vs last week with trend arrowsEach metric shows direction (up/down) and magnitude
Executive summary (5 bullets)150-word email for CEO or VPEach bullet drives a specific decision or action
Your output will vary

The dashboard depends on your pipeline data and campaign metrics. The teaching point is the executive summary: a CEO scanning it on Monday morning should see the headline in 30 seconds -- pipeline health, at-risk deals, forecast status, and the 3 actions that matter this week. No jargon, no dashboards to navigate. Five bullets that drive five decisions.

A Real Morning Workflow in Cowork

There is no standalone daily-briefing skill in this plugin. A rep's morning routine is a short sequence of prompts against the real assets that do exist.

Prompt 1: Check New Signals

Use the `lead-intelligence-agent` to show any new HOT or WARM signals
for NexaFlow accounts since yesterday. Include the recommended next action.

Prompt 2: Check Pipeline Risk

Use the `revenue-reporting-agent` to show this week's at-risk deals,
weighted pipeline, and forecast gap for NexaFlow.

Prompt 3: Prepare for the Top Meeting

Use the `pre-call-brief` skill to prepare me for today's highest-priority
meeting. If it is Meridian Logistics, build the brief for that deal.

If sequence exceptions matter today, prompt the outreach-sequencing-agent next.

Why this is more coherent

This is the real plugin workflow. Cowork invokes named agents and named skills that are actually present in agents/ and skills/. You can save these prompts, reuse them, or chain them into a team routine, but you do not need to invent a new plugin asset to explain the workflow.

Configuring Agent Schedules

Map each agent to the business rhythm. Not every agent runs every day. The schedule reflects when each type of intelligence is most valuable.

In Cowork, this is where /schedule fits. You define the task in natural language, assign the cadence, and Cowork runs that agent task at the chosen time against the same chapter workspace.

CadenceAssetRelated skillsWhy It Matters
MondayRevenue Reporting AgentpipelineStart the week knowing pipeline health and forecast status
WednesdayCRM Hygiene Agentcrm-enrichment + lead-scoringMid-week data quality check catches issues before Friday reporting
FridayMarketing Performance Agentperformance-analysisEnd-of-week campaign review informs next week's budget decisions
DailyLead Intelligence Agentprospect-research + lead-scoringReps need new buying signals before the day starts
ContinuousOutreach Sequencing Agentsequence + outreachEvent-driven -- responds to prospect behaviour in real time
On demandpipeline skillnoneDeep deal review and forecasting before commit calls or QBRs
On demandpre-call-brief skillprospect and deal contextMeeting prep for the deals surfaced by the agents

This schedule means a NexaFlow rep's week looks like this:

Monday morning: Prompt the Revenue Reporting Agent. Know the forecast gap, the at-risk deals, and where leadership attention is required.

Wednesday: CRM Hygiene report arrives. Review flagged records, merge duplicates, update stale data. Ten minutes of maintenance that keeps scoring accurate.

Friday afternoon: Marketing Performance report arrives. See which channels delivered this week. Decide whether to adjust budget for next week before leaving for the weekend.

Every morning: Check the Lead Intelligence Agent. Before any important call, run pre-call-brief. If sequences need attention, prompt the Outreach Sequencing Agent. The rep starts the day informed, not scrambling.

All week: Outreach Sequencing Agent handles touch progression automatically. The rep focuses on conversations, not on remembering which prospect needs touch four.

Example /schedule Prompts

/schedule
Every Monday at 9:00 AM, use the revenue-reporting-agent to build NexaFlow's
weekly revenue dashboard from demo-data.md and my current chapter workspace.
/schedule
Every weekday at 8:00 AM, use the lead-intelligence-agent to check for new HOT
or WARM signals in NexaFlow's target accounts and report the next action.

What You Built

  1. Five RevOps agents understood and configured -- each with a clear purpose, schedule, output format, and skill mapping
  2. Pipeline health audit with deal-level health scores and three-dimension scoring, identifying the 3 highest-risk and 2 strongest deals
  3. Sales forecast with three scenarios (best/likely/worst), revealing a $120,000 gap to quarterly target
  4. A real Cowork routine using lead-intelligence-agent, revenue-reporting-agent, and pre-call-brief instead of a fictional standalone briefing skill
  5. Weekly revenue dashboard with seven metrics and a 150-word executive summary
  6. Agent schedule mapped to NexaFlow's business rhythm -- Monday through Friday, daily and continuous

Try With AI

Use these prompts in your preferred AI assistant.

Prompt 1: Pipeline Analysis

Run /pipeline-review and /forecast on NexaFlow's demo pipeline data
(or your own pipeline if you have one).

From the output, identify:
1. The 3 highest-risk deals and what makes each one risky
2. The 2 most likely to close this quarter and why
3. The gap between likely forecast and quarterly target
4. One deal where you disagree with the agent's probability
estimate -- explain why your assessment differs

What you are learning: Pipeline analysis is not about reading numbers. It is about interpreting deal health in context. The agent scores deals on data it can measure (stage duration, engagement, signals). You evaluate deals on context it cannot measure (relationship strength, procurement cycles, competitive dynamics). The disagreement between your assessment and the agent's is where your judgment adds value.

Prompt 2: Dashboard for a Different Metric

NexaFlow's revenue dashboard tracks 7 metrics focused on pipeline
and acquisition. Reconfigure the Revenue Reporting Agent to produce
a dashboard focused on customer retention instead of pipeline.

Replace the 7 metrics with retention-focused alternatives:
- What metrics would you track? (e.g., churn rate, NPS, expansion
revenue, support ticket volume, renewal pipeline)
- How does the executive summary change when the dashboard measures
retention instead of acquisition?
- What data sources does the agent need that the pipeline dashboard
did not require?

Produce a sample retention dashboard and executive email using
NexaFlow's business context.

What you are learning: Dashboard design is metric selection. Changing from pipeline to retention changes every data source, every threshold, and every recommendation. The underlying agent architecture (schedule-driven, digest-producing, stateless) stays the same. By rebuilding the dashboard for a different business question, you internalise the pattern: agents are configurable instruments, not fixed reports.

Prompt 3: Direct Agent and Skill Workflow

In Cowork, run this sequence for NexaFlow:

1. Use the `lead-intelligence-agent` to show new HOT/WARM signals since yesterday
2. Use the `revenue-reporting-agent` to show at-risk deals, weighted pipeline, and forecast gap
3. Use the `pre-call-brief` skill for the highest-priority meeting today

After reviewing the outputs:
1. Write down the 3 actions you would take today based on
what these outputs told you
2. Now think about what you would have done this morning
WITHOUT these outputs -- would your priorities have been
the same?
3. Identify one signal or risk in the outputs that changes your
plan for today. What would you have missed without it?

If using your own data: which agent or skill output was
most valuable? Which was noise? How would you change the
prompting routine to show more of what matters and less of what does not?

What you are learning: The value of this routine is not information delivery -- it is priority realignment. Without these outputs, you start the day with yesterday's mental model. With them, you start with today's signals, risks, and meeting context. The third question forces you to identify the specific moment where a real plugin asset changed your behaviour. That is the measurable value of automation: not time saved, but decisions improved.

Flashcards Study Aid