Available Context & Tools

@_platform-references/org-variables.md @_platform-references/capabilities.md

Pipeline Focus Task Planner

Goal

Given a list of pipeline deals, produce a single actionable engagement task with a clear, deal-grouped checklist. The core insight is that scattered tasks across many deals lead to shallow engagement with all of them. One focused task with a structured checklist drives deep, intentional engagement with the deals that matter most.

Pipeline Engagement Philosophy

Work the Right Deals, Not All Deals

The natural instinct is to touch every deal a little bit. This is exactly wrong. Research on sales performance reveals:

Reps who focus on their top 3 deals close 34% more revenue than those who spread attention evenly (Gartner Sales Effectiveness Study)
Deal win rates drop 50% when a deal goes 10+ days without meaningful engagement (CSO Insights)
Multi-deal task lists with 8+ items have a 23% completion rate vs. 78% for focused lists of 3-5 items (Asana Work Index)
The average rep has 12-18 open deals but only 3-5 are likely to close in any given period

The Pipeline Focus Task Planner embodies this principle: select the 1-3 deals that deserve your energy right now, then build a single, structured engagement task that ensures deep, productive work on each.

The Single Engagement Task Concept

Instead of creating 10 scattered tasks ("follow up with Acme," "send proposal to TechFlow," "check in with DataBridge"), this skill produces one task with a grouped checklist. Why?

Reduced context-switching: One task to open, one mental model to hold
Visible progress: Checking off sub-items within a single task feels rewarding and trackable
Prioritized flow: The checklist is ordered -- the rep works top to bottom, stopping when time runs out
Clear scope: The task has a due date and a finite checklist. When it's done, it's done.
CRM cleanliness: One engagement task per planning period keeps the task list from exploding

The Alternative (and Why It Fails)

Creating separate tasks per deal sounds logical but causes:

Task list overwhelm (20+ open tasks = decision paralysis)
Cherry-picking easy tasks instead of important ones
Lost context (each task is isolated from the bigger picture)
Incomplete engagement (rep does 1 thing per deal instead of the full engagement motion)

Deal Selection Methodology

See references/deal-selection-rules.md for the complete scoring model with worked examples, velocity benchmarks by deal size, deal rot thresholds, and special selection scenarios.

Primary Selection Criteria

Select up to 3 deals using a weighted scoring model:

DEAL_FOCUS_SCORE = (Urgency x 0.30) + (Risk x 0.25) + (Value x 0.25) + (Momentum x 0.20)

Urgency (30% weight) -- Is there a time-sensitive deadline?

Signal	Score	Rationale
Closes this week	100	Immediate revenue at stake
Closes this month	70	Within active selling window
Close date passed (overdue)	90	Needs immediate stage/date correction
Key meeting scheduled this period	60	Preparation needed
Proposal/contract pending review	80	Decision imminent
No close date set	40	Ambiguity itself is a risk

Risk (25% weight) -- Is the deal in danger?

Use deal_health_scores table for risk assessment:

Signal	Score	Rationale	Health Score Field
Health score < 30	100	Critical -- intervention required	`overall_health_score < 30`
Health score 30-50	75	At risk -- needs this-week attention	`overall_health_score 30-50`
Champion went dark (7+ days)	90	Single-threaded deal losing its thread	Check `relationship_health_scores.is_ghost_risk` or `days_since_last_contact > 7`
Risk level = critical	95	Pre-computed critical risk	`deal_health_scores.risk_level = 'critical'`
Risk level = high	80	Pre-computed high risk	`deal_health_scores.risk_level = 'high'`
Sentiment declining	75	Buyer engagement deteriorating	`deal_health_scores.sentiment_trend = 'declining'`
Competitor mentioned in last meeting	70	Evaluation risk	(manual detection from notes)
Stakeholder change (new decision-maker)	65	Relationship restart needed	(manual detection)
No activity in 10+ days	80	Deal is stalling	`deal_health_scores.days_since_last_activity > 10`

Health score integration: Query deal_health_scores and relationship_health_scores for each deal in pipeline. Use overall_health_score and risk_level as primary risk scoring inputs. Use risk_factors array to identify specific interventions needed.

Value (25% weight) -- Is it worth the focus?

Deal Value Relative to Average	Score
3x+ portfolio average	100
2-3x average	75
1-2x average	50
Below average	25

Momentum (20% weight) -- Can we make progress this period?

Signal	Score	Rationale
Next step is clearly defined	80	Actionable engagement possible
Stakeholder recently replied	70	Engagement window is open
Meeting on calendar this period	90	Natural touchpoint exists
Waiting on external dependency	20	Low leverage -- action may be blocked
Stage recently advanced	60	Positive trajectory to maintain

Selection Rules

Always 1-3 deals: Never 0 (even a clean pipeline has optimization opportunities), never more than 3 (the whole point is focus).
At least 1 high-urgency deal if any exist (closing soon or at risk).
Diversity of stage: Avoid selecting 3 deals all at the same stage. Mix stages for varied engagement types.
No deals updated today: If a deal was already meaningfully engaged today, it does not need focus task attention.
Respect capacity: If user is "busy," select 1 deal. If "normal," select 2-3. If "available," select 3.

When You Cannot Select 3

Scenario	Response
Fewer than 3 deals in pipeline	Select all. Note: "Small pipeline -- consider prospecting."
All deals are healthy and active	Select the 1-2 with the nearest close dates. Focus on acceleration, not rescue.
All deals are stale / at risk	Select the top 3 by value. Triage mode -- focus on saving the most revenue.

Checklist Design Principles

The checklist is the core deliverable. Each item must be. Consult references/task-templates.md for complete task templates by deal stage and activity type, including description templates, duration estimates, success criteria, and checklists.

Specific

Not: "Follow up with Sarah" Yes: "Email Sarah Chen the updated pricing sheet with the volume discount she requested on Tuesday's call"

Actionable

Not: "Think about next steps for TechFlow" Yes: "Draft 3 discovery questions for Mike Ross focused on their Q2 budget timeline"

Time-Estimated

Every checklist item includes a time estimate in parentheses. This helps the rep plan their day and know when to stop.

Format: - [ ] [Action description] (~X min)

Ordered by Impact

Within each deal group, checklist items are ordered:

The action most likely to advance the deal stage
The action that addresses the biggest risk
Supportive actions (logging, updating, preparing)

Outcome-Oriented

Each item should have an implicit "done when" condition. The rep should know exactly when to check the box.

Not: "Work on the Acme proposal" Yes: "Send the customized Acme proposal PDF to Sarah Chen via email (~15 min)"

Checklist Structure by Deal

The task description groups checklist items under deal headings:

## Deal 1: [Deal Name] ($[Value] - [Stage])
Why now: [One-line rationale]

- [ ] [Primary action -- highest impact] (~X min)
- [ ] [Secondary action -- risk mitigation] (~X min)
- [ ] [Support action -- log/update/prepare] (~X min)

## Deal 2: [Deal Name] ($[Value] - [Stage])
Why now: [One-line rationale]

- [ ] [Primary action] (~X min)
- [ ] [Secondary action] (~X min)
- [ ] [Support action] (~X min)

## Deal 3: [Deal Name] ($[Value] - [Stage])
Why now: [One-line rationale]

- [ ] [Primary action] (~X min)
- [ ] [Secondary action] (~X min)

---
Total estimated time: ~X min

Items Per Deal

Critical deal (health < 50 or closing this week): 3-4 items
High-priority deal: 2-3 items
Monitoring deal: 1-2 items
Never exceed 4 items per deal -- if more is needed, the deal needs a dedicated session, not a checklist item

Total Checklist Size

Busy capacity: 3-5 items total across 1 deal
Normal capacity: 6-9 items total across 2-3 deals
Available capacity: 9-12 items total across 3 deals

Period-Based Planning

The planning period affects deal selection and checklist depth:

This Week

Focus on deals closing this week or next
Checklist items are immediate actions (today/tomorrow)
Task due date: End of current week (Friday)
Emphasis: Close, advance, or save deals in the near-term window

This Month

Include deals closing this month and next
Checklist items include multi-step sequences (first contact, follow-up, close)
Task due date: End of current month or 2 weeks from now (whichever is sooner)
Emphasis: Stage advancement and pipeline progression

This Quarter

Strategic view across the full pipeline
Checklist items are milestone-oriented (get to Proposal stage, schedule exec meeting)
Task due date: End of current month (review monthly, not quarterly)
Emphasis: Pipeline building and long-term positioning

Capacity Adjustment Rules

Capacity	Deals Selected	Items Per Deal	Total Items	Task Complexity
Busy	1	3-4	3-5	Simple, quick-win actions only
Normal	2-3	2-3	6-9	Balanced mix of quick and deep actions
Available	3	3-4	9-12	Include prep work and strategic actions

Auto-detection (when capacity is not specified):

Check today's meeting count: 5+ = busy, 2-4 = normal, 0-1 = available
Check open task count: 15+ open = busy, 5-14 = normal, 0-4 = available
When signals conflict, default to "normal"

Required Capabilities

CRM: To fetch deal pipeline status, health scores, and activity history

Inputs

pipeline_deals: Output from execute_action("get_pipeline_deals", { filter: "closing_soon", period: "this_week", include_health: true, limit: 10 }) -- should include deals and health if available
Health scores: Query deal_health_scores table for all pipeline deals to get overall_health_score, risk_level, risk_factors, sentiment_trend, days_in_current_stage, days_since_last_activity
Relationship health: Query relationship_health_scores for primary contacts to identify ghost risk and response patterns
period (optional): "this_week" | "this_month" | "this_quarter" -- defaults to "this_week"
user_capacity (optional): "busy" | "normal" | "available" -- defaults to "normal"

Prioritization with health data: Use combined health scores (deal + relationship) to weight the focus score calculation. Deals with low health scores but high urgency/value get rescue-focused checklists. Deals with high health scores get acceleration-focused checklists.

Output Contract

Return a SkillResult with:

data.task: The single engagement task
- title: string -- short, action-oriented (e.g., "Pipeline Focus: Engage Top 3 Deals This Week")
- description: string -- includes the full grouped checklist (see Checklist Structure above)
- due_date: ISO date string (default: end of current week / Friday)
- priority: "low" | "medium" | "high" (high if any selected deal is critical)
- estimated_minutes: number -- sum of all checklist item estimates
- deal_count: number -- how many deals are covered
data.top_deals: array of up to 3 selected deals
- id: string
- name: string
- value: number
- stage: string (current deal stage)
- health_score: number | null
- days_stale: number
- close_date: string | null
- why_now: string (one-sentence rationale for selection)
- focus_score: number (0-100, transparency into selection)
data.rationale: string -- 2-3 sentence explanation of why these deals were chosen over others. Include the selection methodology and key signals.
data.excluded_deals: array (optional, for transparency)
- name: string
- reason: string (why not selected, e.g., "healthy and recently engaged" or "blocked by external dependency")

Rationale Communication

The rationale must explain why these deals, not those. Users should understand and trust the selection.

Good Rationale Example

"Selected Acme Corp ($45K, closing Friday), TechFlow ($28K, champion dark 9 days), and DataBridge ($62K, contract pending). Acme is your nearest close and needs a pricing follow-up before the weekend. TechFlow's champion silence is becoming critical -- re-engagement this week could save a $28K deal. DataBridge has the highest value and a contract review that's been waiting 3 days. Together, these 3 deals represent $135K in pipeline at risk of stalling without this week's engagement."

Bad Rationale Example

"These are your top 3 deals that need attention." -- This tells the user nothing they did not already know.

Rationale Structure

Name the deals with values and key signals
Explain each selection in one clause
Quantify the aggregate (total pipeline value at stake)
Connect to the period (why this week specifically)

Quality Checklist

Before returning the engagement task, verify:

Error Handling

Empty pipeline

Return a task focused on pipeline building:

Title: "Pipeline Building: Prospecting and Outreach"
Checklist: prospecting actions, old lead re-engagement, network outreach
Rationale: "No active deals in pipeline. Focus shifts to pipeline generation."
Do NOT return an empty result -- always provide actionable guidance.

All deals are healthy and recently engaged

This is a positive signal. Return an acceleration-focused task:

Select the top 1-2 deals by close date
Checklist items focus on advancing stage, not rescuing
Rationale: "Pipeline is healthy. Focus on accelerating your nearest close opportunities."

No health scores available

Calculate selection using urgency, value, and staleness only (drop the health component and redistribute its 25% weight to urgency and staleness). Note: "Deal health scores unavailable -- selection based on close date, value, and activity recency. Recommend running health recalculation to enable health-informed prioritization."

Single deal in pipeline

Select it. Build a thorough checklist (4-5 items) since all focus goes to one deal. Add a prospecting item at the end: "Identify 2-3 new prospects to diversify pipeline."

Deals with missing close dates

Penalize in urgency scoring (score 40 instead of higher values) but do not exclude. Add a checklist item: "Confirm close date with [contact] and update CRM."

Period mismatch (no deals closing in the selected period)

Expand the window. If "this_week" returns no closing deals, look at "this_month." If "this_month" returns none, look at the full pipeline. Note the expansion: "No deals closing this week. Expanded to this month's pipeline."

Very large pipeline (20+ deals)

Pre-filter to the top 10 by the scoring model before applying the 1-3 selection. Never process more than 10 deals in detail -- it adds latency without improving selection quality.

Conflicting signals (high value but very stale)

When a deal scores high on value but also very high on staleness (14+ days), it may be dead. Add a diagnostic checklist item first: "Assess deal viability -- send a 'still interested?' check-in before investing more time." If the deal is truly dead, the rep should disqualify it, not work it.