ai-native-workflow
Jeff Su's AI-Native Workflow framework (4 Habits) mapped onto Nebuchadnezzar v3.4 pipeline infrastructure. Use when:(1) "swipe file" or "what worked before" - retrieve proven email patterns by stage(2) "task mapping" or "M/AI split" or "YOU vs AI" - get decisions for stage transitions(3) "breadcrumb" or "anchor this" - link conversations to HubSpot + Git deal folders(4) "prompt library" or "battle-tested" - access working slash commands(5) "Blue Pill" - customer-facing communication (emails, proposals)(6) "Red Pill" - internal analysis (PLD, Brand Scout, Deal Snapshot)(7) Stage transitions [00→01], [02→03], [04] follow-ups, [08→09] win-backsTriggers on: "ai-native", "4 habits", "swipe", "winning email", "what worked", "task mapping", "YOU vs AI", "blue pill", "red pill"
When & Why to Use This Skill
This Claude skill implements Jeff Su's '4 Habits' AI-Native Workflow framework integrated with the Nebuchadnezzar v3.4 pipeline infrastructure. It optimizes sales operations by providing a structured system for retrieving proven email patterns, mapping tasks between human and AI execution, and ensuring all conversation data is anchored to HubSpot and Git. By utilizing 'Blue Pill' (external) and 'Red Pill' (internal) modes, it streamlines the entire sales cycle from lead generation to win-back strategies, enhancing productivity and CRM data integrity.
Use Cases
- Sales Outreach Optimization: Retrieve and adapt 'swipe file' email patterns for specific pipeline stages, such as cold outreach [00→01] or proposal follow-ups [04].
- Task Mapping (YOU vs AI): Decide which workflow steps require human intuition (YOU) and which can be automated (AI) using a granular task-template grid.
- CRM Data Anchoring: Automatically link AI-generated meeting debriefs, email drafts, and analysis reports to HubSpot notes and Git deal folders using the '10-minute rule'.
- Internal Deal Analysis: Use 'Red Pill' mode to generate internal-only artifacts like Brand Scout lead intelligence, PLD shipping analysis, and pipeline snapshots.
- Command-Driven Pipeline Management: Access a battle-tested prompt library to execute specific sales actions like /brand-scout for research or /log-activity for CRM updates.
| name | ai-native-workflow |
|---|---|
| description | | |
| Triggers on | "ai-native", "4 habits", "swipe", "winning email", "what worked", "task mapping", "YOU vs AI", "blue pill", "red pill" |
AI-Native Workflow Skill
"I can only show you the door. You're the one that has to walk through it."
Overview
Transform pipeline operations into AI-native workflows using Jeff Su's 4 Habits framework, fully integrated with the Nebuchadnezzar v3.4 infrastructure.
Core Principle: AI drafts, you decide timing and personal touches.
The 4 Habits
| Habit | Purpose | Integration Point |
|---|---|---|
| AI Breadcrumbs | Hyperlink AI chats to workspace | HubSpot Notes + Git _NOTES.md |
| AI Swipe Files | Provide examples of excellent work | references/swipe-files/ by stage + mode |
| AI-First Task Planning | Map tasks as Manual/AI before projects | references/task-templates.md |
| Prompts Database | Save working prompts by use case | references/prompt-library.md |
Mode System: Blue Pill vs Red Pill
Blue Pill Mode (Customer-Facing)
External communication that goes to prospects/customers.
| Stage | Swipe File | Content Type |
|---|---|---|
| [00→01] | discovery-outreach.md |
Cold outreach, follow-up cadence |
| [04] | proposal-followups.md |
Day 1/3/7/10/14 follow-ups |
| [08→09] | winbacks.md |
Re-engagement emails |
| All | patterns-summary.md |
Golden rules, quick diagnosis |
Red Pill Mode (Internal Analysis)
Internal work products that stay in the system.
| Analysis Type | Swipe File | Output |
|---|---|---|
| PLD Analysis | internal-analysis.md |
Customer Shipping Analysis |
| Transit Report | internal-analysis.md |
Transit Performance Summary |
| Brand Scout | internal-analysis.md |
Lead Intelligence Report |
| Discovery Prep | internal-analysis.md |
Call Prep Sheet |
| Meeting Debrief | internal-analysis.md |
HubSpot Note |
| Deal Snapshot | internal-analysis.md |
Pipeline Position |
Workflows
Swipe File Lookup
When user asks for "winning email" or "what worked" for a stage:
- Identify pipeline stage ([00], [02], [04], [09], etc.)
- Determine mode: Blue Pill (external) or Red Pill (internal)
- Load appropriate swipe file from
references/swipe-files/ - Present pattern + verbatim example
- Adapt to current prospect context
Quick Routing:
"Need to reach out cold" → discovery-outreach.md
"Proposal follow-up" → proposal-followups.md
"Win them back" → winbacks.md
"Analyze this data" → internal-analysis.md
"What's the pattern?" → patterns-summary.md
Task Mapping (YOU vs AI)
When user needs M/AI decisions for a task:
- Identify stage transition or activity type
- Load
references/task-templates.md - Present YOU vs AI grid with granular breakdown
- Suggest specific tools/commands for AI tasks
- Identify swipe file for reference
Example Output:
STAGE: [02]→[03] Discovery Complete → Rate Creation
YOU TASKS:
- Review discovery notes
- Submit rate request
- Follow up with pricing
AI TASKS:
- Draft data request email → /create-followup
- Run shipping analysis → analysis_skill.md
- Generate Customer Shipping Analysis → internal-analysis.md
SWIPE FILE: internal-analysis.md (PLD Analysis pattern)
Breadcrumb Anchoring
When user says "anchor this" or conversation produced reusable output:
- Primary anchor: HubSpot deal notes via
/log-activity - Secondary anchor: Git deal folder
_NOTES.mdorCustomer_Relationship_Documentation.md - Cross-link: Include "Full analysis: See [Deal]_Folder/_NOTES.md" in HubSpot
10-Minute Rule: If conversation >10 minutes OR produces reusable output → anchor immediately.
Anchor Destinations by Output Type:
| Output | Primary | Secondary |
|---|---|---|
| Email draft | — (it gets sent) | — |
| Meeting debrief | HubSpot Note | Deal folder |
| PLD analysis | Deal folder | HubSpot summary |
| Brand Scout | Brand Scout Reports/ |
HubSpot company |
| Deal snapshot | HubSpot Note | — |
Prompt Lookup
When user asks "what command for X" or needs a battle-tested prompt:
- Check
references/prompt-library.mdfor command - Provide usage syntax and reliability rating
- Match command to current stage
- Suggest alternatives if primary doesn't fit
Reference Files
| File | Load When | Mode |
|---|---|---|
references/swipe-files/discovery-outreach.md |
[00→01] cold outreach | Blue Pill |
references/swipe-files/proposal-followups.md |
[04] email follow-ups | Blue Pill |
references/swipe-files/winbacks.md |
[08→09] re-engagement | Blue Pill |
references/swipe-files/patterns-summary.md |
General "what works" | Both |
references/swipe-files/internal-analysis.md |
PLD, Brand Scout, Snapshots | Red Pill |
references/task-templates.md |
Task planning, YOU vs AI | Both |
references/prompt-library.md |
Command lookup | Both |
Quick Reference
Swipe File by Stage
| Stage | Blue Pill (External) | Red Pill (Internal) |
|---|---|---|
| [00] Lead Gen | discovery-outreach.md |
Brand Scout reports |
| [01] Discovery Scheduled | discovery-outreach.md |
Discovery Prep Sheet |
| [02] Discovery Complete | patterns-summary.md |
Meeting Debrief |
| [03] Rate Creation | patterns-summary.md |
Customer Shipping Analysis |
| [04] Proposal Sent | proposal-followups.md |
Deal Snapshot |
| [05] Setup Docs | patterns-summary.md |
— |
| [06] Implementation | — | — |
| [07] Closed Won | — | — |
| [08] Closed Lost | winbacks.md |
Loss analysis |
| [09] Win-Back | winbacks.md |
— |
Core Commands (High Reliability)
/brand-scout:scout [Company] # Lead research
/check-lead [Company] # HubSpot duplicate check
/add-lead [Company] # Create company + contacts
/cvm-goals weekly # Pipeline dashboard
/check-tasks # Daily task review
python unified_sync.py 9am # Morning sync
python unified_sync.py eod # End-of-day sync
Support Commands (Medium Reliability)
/create-followup # Draft follow-up email
/deal-snapshot [Deal] # Quick status
/update-deal # Stage change
/log-activity # Anchor breadcrumb
/meeting-summary # Summarize call
/pipeline-health # Stage analysis
Integration Points
HubSpot Stage IDs
Discovery Scheduled: 1090865183
Discovery Complete: d2a08d6f-cc04-4423-9215-594fe682e538
Rate Creation: e1c4321e-afb6-4b29-97d4-2b2425488535
Proposal Sent: d607df25-2c6d-4a5d-9835-6ed1e4f4020a
Setup Docs: 4e549d01-674b-4b31-8a90-91ec03122715
Implementation: 08d9c411-5e1b-487b-8732-9c2bcbbd0307
Closed Won: 3fd46d94-78b4-452b-8704-62a338a210fb
Closed Lost: 02d8a1d7-d0b3-41d9-adc6-44ab768a61b8
File Paths
Deal folders: [XX-STAGE]_Deal_Name/
Brand Scout: Brand Scout Reports/
Templates: Communication-Systems/Templates/
Daily log: _DAILY_LOG.md
Sync reports: sync_reports/
Existing Skills (Reference, Don't Duplicate)
| Skill | Purpose | When to Use Instead |
|---|---|---|
brand_scout_skill.md |
Lead research | Deep company research |
cvm-white-glove-sales-process |
Stages 02-07 process | 38 Questions framework |
december-pipeline-priming |
Q1 campaign templates | GRI-timed outreach |
analysis_skill.md |
PLD/shipping analysis | Full 9-tab Excel reports |
nate-jones-deliberate-practice |
Artifact evaluation | Self-scoring |
Daily Workflow Integration
Morning (9am):
1. Run sync: python unified_sync.py 9am
2. Review pipeline health
3. Pick 3 priority actions (YOU decision)
4. Load task-templates.md for YOU vs AI split
5. Execute micro-actions
Throughout Day:
- Before email: Check swipe file for stage
- After call: Anchor meeting debrief (10-min rule)
- When stuck: Check prompt-library.md
End of Day:
1. Run sync: python unified_sync.py eod
2. Anchor any unanchored outputs
3. Clear or reschedule remaining tasks
Anti-Patterns
NEVER:
- Send AI drafts without review
- Skip checking swipe file before writing
- Invent prompts when battle-tested ones exist
- Let AI outputs disappear (10-minute rule)
- Use Blue Pill patterns for internal analysis
- Use Red Pill patterns for customer emails
ALWAYS:
- Check swipe file BEFORE drafting
- Review and personalize AI drafts
- Anchor outputs within 10 minutes
- Map YOU vs AI before complex tasks
- Match command to current stage
Skill Version: 2.0 Last Updated: December 2025 Framework: Jeff Su's 4 Habits + Nebuchadnezzar v3.3