Pipeline Reconciliation Tracker

Living document reconciling all workstreams. Read first, update after each build pass.

Workstream 1
MAP
→
Workstream 2
RADAR
→
Workstream 3
ENGINE
→
Workstream 4
FACTORY
→
Workstream 5
PRODUCT

Cross-Taxonomy Reconciliation 3 systems → 1 map

Three documents describe the extraction system with different vocabularies. This table is the canonical mapping. When any source changes, update here first.

Source Documents

ViewDocumentLocationDescribes
SwarmAgentic Boarding PRD v1.0expert.align360.io + E:\align360\_extraction-files\18 color-coded agents, 4-pass protocol, external/partner-facing
SkillsRecursive Clone Architecture v2expert.align360.io18 skills, 8 layers (L0→L4), build spec
ModulesMaster Registry PRD 4plan.jasondmacdonald.com8+1 extraction modules, structured JSON output spec

Agent ↔ Skill ↔ Module Mapping

Swarm Agent(s)Skill (v2)PRD 4 ModuleStatus
Chief of Staffboarding-orchestrator (L4)OrchestrationBuilt — needs 24-card update
PURPLE × 7 (positioning)deep-research + expert-recon (L0)Gate 2 intel + PRD 5Built (2 skills cover 7 agents)
GREEN (brand extraction)design-system-extractorUtilityBuilt
BLUE: Transcript Analyzersoul-extractor (L2)Mod 1 (Thinking) + Mod 7 (Pattern)Built
BLUE: Document Extractorframework-extractor (L2)Mod 4 (IP) + Mod 5 (Modular) + Mod 6 (Meta)Built — needs +micro-units
BLUE: Content Auditorresource-extractor (L2)Mod 8 (Templates)Built
—voice-extractor (L2)Mod 2 (Voice)Built — needs +pattern-breaking
—expert-framework-creator (L0.25)Mod 1 (primary)UNBUILT — CRITICAL DOMINO
PURPLE: Psychology Agentsoul-extractor (partial)Mod 3 (CTA Psychology)Partial
—offer-extractor (L2)Mod 3 (monetization)UNBUILT
—rubric-builder (L1)Pass 3 (Rubric Building)Built — needs dual validation
—gap-analyzer (L2.5)Pass 4 (Gap Analysis)Built
Builder: SysPrompt + KnowledgeFileclone-compiler (L3)Mod 8 + deploymentBuilt — needs +user profile +offers
Builder: Lead Magnetlead-magnet-builder (L3)PRD 7Built
Builder: Onboarding Flowonboarding-builder (L3)PRD 7Built
Builder: Offer Pageno skillPRD 9GAP — no dedicated skill
—clone-tester (L3.5)3-pass validationUNBUILT
—demo-compiler (L0.5)Pre-deal artifactUNBUILT
YELLOW (network/LinkedIn)no skillNo moduleGAP — deprioritized
RED (domain generation)no skillNo moduleGAP — GoDaddy MCP exists
—masterybook-sync (L0)UtilityBuilt
——Mod 9 (Retrieval Patterns)GAP — no skill, no agent

4-Pass Protocol ↔ Skill Layers

Pass (PRD v1.0)Skill Layer (v2)PRD 4 Module
Pass 1: Raw ExtractionL2 extractors (soul, voice, framework, resource)Modules 1-8
Pass 2: Pattern Recognitionexpert-framework-creator (L0.25) + cross-refModule 1 primary
Pass 3: Rubric Buildingrubric-builder (L1)Quality gates
Pass 4: Gap Analysisgap-analyzer (L2.5)Completeness audit

Terminology Collisions

TermBoarding Pack MeaningMaster Registry MeaningResolution
GatesOnboarding milestones (Brand Assets → Offer Approval)Data scoping tiers (Operator → Competitor)Rename boarding gates to "Boarding Milestones"
ModulesNot used8+1 extraction content typesAdopt universally for extraction content
SkillsNot usedClaude Code SKILL.md filesAdopt universally for executable tools

Current Pipeline Status as of 2026-03-26

FACTORY Layer: Skills Inventory

SkillLayerFile LocationLast ModifiedStatus
deep-researchL0~/.claude/skills/deep-research/SKILL.md2026-03-10Built
expert-reconL0~/.claude/skills/expert-recon/SKILL.md2026-03-10Built
masterybook-syncL0~/.claude/skills/masterybook-sync/SKILL.md2026-03-10Built
soul-extractorL2~/.claude/skills/soul-extractor/SKILL.md2026-03-09Built
voice-extractorL2~/.claude/skills/voice-extractor/SKILL.md2026-03-09Needs +pattern-breaking
framework-extractorL2~/.claude/skills/framework-extractor/SKILL.md2026-03-09Needs +micro-units
resource-extractorL2~/.claude/skills/resource-extractor/SKILL.md2026-03-09Built
rubric-builderL1~/.claude/skills/rubric-builder/SKILL.md2026-03-09Needs dual validation
gap-analyzerL2.5~/.claude/skills/gap-analyzer/SKILL.md2026-03-09Built
clone-compilerL3~/.claude/skills/clone-compiler/SKILL.md2026-03-09Needs +user profile +offers
lead-magnet-builderL3~/.claude/skills/lead-magnet-builder/SKILL.md2026-03-09Built
onboarding-builderL3~/.claude/skills/onboarding-builder/SKILL.md2026-03-09Built
boarding-orchestratorL4~/.claude/skills/boarding-orchestrator/SKILL.md2026-03-10Needs 24-card + modes
design-system-extractorUtility~/.claude/skills/design-system-extractor/SKILL.md2026-03-09Built
expert-framework-creatorL0.25——UNBUILT
demo-compilerL0.5——UNBUILT
offer-extractorL2——UNBUILT
clone-testerL3.5——UNBUILT

PRODUCT Layer: Samuel Ngu / Align360

PRD 4 ModuleSource MaterialExtracted?Structured JSON?
Mod 1: Thinking StructuresFLC Framework, Tri-Filter, Clarity PathPartial (buried in v6.1)No
Mod 2: Voice & StyleBoarding pack ("pastoral but direct, 8th grade")YesNo
Mod 3: CTA PsychologyMonetization guardrails, Laws of BabylonPartialNo
Mod 4: Embedded IP36 stacks, 9 Wiring Types, frameworksYesPartial (tables exist)
Mod 5: ModularizationPhase 0-4 progression, stack sequencingYesNo (list format)
Mod 6: Meta-Structures5-phase architecture, onboarding flowYesNo
Mod 7: Pattern RecognitionWiring Types, RhythmOS, Decision ContextYesNo
Mod 8: Prompt TemplatesExtraction transcripts with expert responsesRaw material existsNo
Mod 9: Retrieval Patterns14 background systemsYes (descriptive)No

Key finding: Content is gathered and documented. The gap is transformation into structured JSON, not research.

Temporal Log newest first

2026-03-26 | SESSION: full-reconciliation-review
Full Pipeline Reconciliation — 6 Source Documents Reviewed

Reviewed and cross-referenced all published planning/architecture documents to establish canonical pipeline state.

  • Discovery: Three taxonomies (Swarm/Skills/Modules) describe the same extraction system with different vocabularies. Canonical mapping created in this tracker.
  • Discovery: "Gates" terminology collision between boarding pack (onboarding milestones) and master registry (data scoping tiers). Resolution: rename boarding gates to "Boarding Milestones."
  • Discovery: expert.align360.io/index returns 404. Needs fix or redirect to /index-new.
  • Discovery: boarding-architecture (v1, 14 skills) is explicitly superseded by recursive-clone-v2 (v2, 18 skills). Should add redirect banner.
  • Discovery: No skill files modified since 2026-03-10. Spring break pause (Mar 17-23). Team resumed Mar 24.
  • Discovery: PRD 4 Module 1 (Thinking Structures / FLC Framework) confirmed as critical domino. expert-framework-creator skill (L0.25) is the tool that executes it. Currently unbuilt.
  • Discovery: Samuel pressing for testing access (WhatsApp to Sumit). Will + Sumit need to finalize betaapp.io.
  • Vision shift (Mar 24-25 calls): Align360 evolving from structured GUI → agent-first dynamic experience. Extraction pipeline feeds agent brain, not just chat system prompt.
  • GTM pivot: Labs deprioritized → implementation partner vehicle for MasteryOS. Clone messaging: "10x monetization" not "own your IP."

Second-order effects of this reconciliation:

  • Any new session can now read this tracker + the MAP change queue and know exactly where to pick up
  • Skill builds can validate against all 3 taxonomies simultaneously
  • MAP updates can be batched rather than one-off forgotten changes
  • Pipeline visual (MAP→RADAR→ENGINE→FACTORY→PRODUCT) is now the lingua franca across all docs
Master Registry A360 Reconciliation Execution Playbook Clone Architecture v2 Swarm PRD Command Center Boarding Pack
2026-03-25 | SESSION: will-derek-aaron-call
Team Call — Labs Pivot, GTM, Agent-First Vision
  • Monetization Lab deprioritized → implementation partner for MasteryOS
  • Aaron Swensen joining as implementation/content builder
  • Tony Gaskins meeting in ~1 month (potential clone client)
  • Jason: "By mid-April, 2 operating systems running (human + agent), then reconcile"
  • Jason: "Finishing Align360 as a skill, so that it's agentic"
  • Friday deadline: defined GTM / implementation vehicle
Transcript: 26.3.25 will derek aaron jason.txt
2026-03-24 | SESSION: sumit-dev-standup
Dev Standup — Agent Branch Strategy, Samuel ETA
  • Samuel asking for testing access. Will needs to finalize prompts/content on betaapp.io
  • Hidden admin subscription plan needed (test features before public exposure)
  • Agent coding branch: Sumit to architect bot-vs-human branch + merge strategy
  • Goal: Claude Code doing all coding by mid-April, human devs for last 10%
  • Key Vault integration in progress (Rohit), Stripe webhooks updated (Ashwani)
  • Cross-tenant login bug: FreedomLife users can log into Align360 (needs filter)
  • Vision discussion: Align360 as AGENT, not structured UI. Dynamic experience generation.
Transcript: 26.3.4 sumit call.txt
2026-03-10 | SESSION: skill-build-sprint
14 Skills Built — Sprint Freeze Before Spring Break
  • All 14 skills finalized and saved to ~/.claude/skills/
  • boarding-orchestrator, deep-research, expert-recon, masterybook-sync updated last
  • Samuel Layer 0 recon complete: intel report published to /samuel-ngu-intel
  • Wiring for Impact lead magnet published
  • v2 architecture published to /recursive-clone-v2
  • 4 skills remaining: expert-framework-creator, demo-compiler, offer-extractor, clone-tester

MAP Change Queue batch updates needed

Changes discovered during builds that need to be propagated back to planning documents. Process: collect here → batch update → mark complete.

master-registry
Add cross-taxonomy mapping table (Swarm ↔ Skills ↔ Modules) to Section 06a or new Section 17
High
master-registry
Update Section 12 (Skills Inventory) — 14 skills have SKILL.md files, not zero MANUAL.md. Reconcile MANUAL.md convention with SKILL.md reality
High
align360-reconciliation
Rename "Gates" to "Boarding Milestones" throughout to eliminate terminology collision with master registry Gate system
Medium
execution-playbook
Add step for publishing temporal log entry after each execution step (three-artifact rule: customer artifact + reusable skill + architecture validation + temporal log entry)
Medium
recursive-clone-v2
Add Module 9 (Retrieval Patterns) to skill architecture — currently has no skill or agent mapping
Medium
agentic-boarding-prd
Add note: PURPLE 7-agent swarm is aspirational; current implementation uses 2 skills (deep-research + expert-recon). Consider if granular agents are needed or if 2 skills suffice.
Low
expert.align360.io/index
Currently 404. Either redirect to /index-new or republish
Medium
boarding-architecture
Add banner: "Superseded by recursive-clone-v2" with link
Low
plan index
Add link to this pipeline tracker + temporal log. Make it the canonical "where are we" document.
High

Second-Order Effects cascading impacts

Effect of building expert-framework-creator (the domino)
Validates PRD 4 Module 1 → creates reusable extraction skill → enables rubric-builder expansion (dual validation) → enables all downstream extractors to output structured JSON → enables clone-tester to score against expert's own framework → changes the quality bar for every future expert clone
Effect of structured JSON output (Modules 1-9)
Monolithic System Prompt v6.1 decomposes into modular, retrievable units → enables agent-first dynamic experience (no static sidebar) → enables personalized tool routing per user → enables cross-expert framework comparison → enables the "download Samuel" paradigm from Sumit call vision
Effect of this temporal log existing
Any new Claude session reads this page and knows: where we are, what changed, what to build next, and what MAP changes are queued → eliminates session-to-session knowledge loss → enables concurrent builds without drift → enables batch MAP updates instead of forgotten one-offs
Effect of three-taxonomy mapping
Skill builds validate against all 3 source docs simultaneously → prevents building something that contradicts another doc → enables gap identification (Mod 9, offer page builder, YELLOW/RED agents) → enables informed deprioritization vs. accidental omission

Session Handoff Protocol read this first

For any new Claude Code session working on the extraction pipeline:

1. Read These Files (in order)

#FilePurpose
1This page (pipeline-reconciliation.html)Where we are, what maps to what, what's queued
2~/.claude/projects/E--align360/memory/MEMORY.mdProject memory, team, architecture overview
3~/.claude/projects/E--align360/memory/reconciliation.mdCompact reconciliation state for context window
4Relevant SKILL.md file for the skill being built/expandedCurrent skill spec

2. After Each Build Pass

ActionWhere
Add temporal log entryThis page → #temporal-log section
Update skill statusThis page → #status section
Queue any MAP changes discoveredThis page → #change-queue section
Update memory/reconciliation.mdLocal memory file
Republish this pageplan.jasondmacdonald.com/pipeline-tracker

3. Build Priority (Current)

PriorityActionWhy
1Build expert-framework-creator (L0.25)Critical domino — Module 1 extraction, everything cascades
2Run Module 1 on Samuel's materialValidates pipeline + produces customer artifact
3Expand extractors for structured JSON outputTransforms gathered content into modular units
4Build offer-extractor (L2)Covers Module 3 + "10x monetization" positioning
5Build clone-tester (L3.5)Closes quality loop with 85%+ threshold
6Build demo-compiler (L0.5)Non-blocking, for JV sales (Tony Gaskins meeting)