MOC

Auset Brain — Map of Content

The platform’s persistent second brain. Knowledge is never lost, never erased, always accessible. Named after Carter G. Woodson — the Father of Black History.

Vault Structure

auset-brain/
├── Daily/          # Session logs, daily notes
├── Projects/       # Per-Heru context, products, ecosystem
├── Agents/         # One note per named agent
├── Decisions/      # Architecture, infrastructure, protocols
├── Feedback/       # Corrections from Amen Ra (NEVER lose these)
├── People/         # Team, clients, stakeholders, credentials
├── Canvas/         # Visual relationship maps
└── Templates/      # Note templates

---

LESSONS LEARNED (read every session)

• three-tier-agent-architecture — Opus does NOT dispatch or write code; Haiku dispatches, agents write code
• feedback-verify-before-declaring-done — Tester bugs = immediate priority (within MINUTES)
• quiknation-platform-vision — WCR client demo failure: client meeting = #1 immediately
• amplify-env-var-safety — Amplify env var wipe: ALWAYS read-modify-write
• feedback-never-ask-quik-to-run-commands — Never ask Quik to run commands; he’s a tester, not a developer
• feedback-never-touch-screen-sharing — Never touch Screen Sharing; breaks Quik Cloud
• feedback-never-show-quik-cloud — Never show Quik Cloud publicly; CLASSIFIED
• feedback-slack-plain-language — Slack = plain language only in maat-agents
• feedback-apple-wwdr-g3-required — Apple WWDR G3 cert required on every build Mac
• feedback-qc1-keychain-partition-list — QC1 keychain partition list: unlock + set before EAS builds
• feedback-pgcmc-not-priority — PGCMC NOT a priority; removed from pipeline
• feedback-cursor-is-primary-agent — Cursor is PRIMARY agent; 1 Haiku per farm, Cursor does ALL coding
• feedback-no-bedrock-agents — NEVER use Bedrock for agents; use Cursor Agent CLI
• feedback-check-memory-before-acting — Check memory before acting on infra; don’t explore deprecated systems
• feedback-read-memory-before-acting — CHECK MEMORY BEFORE EVERY ACTION

---

People (10 notes)

Amen Ra — CTO & Co-Founder

• user-quiknation-email — Official email and title
• user-origin-story — Sept 18, 2018 first call with Quik
• user-origin-with-ai — AI journey
• user-family-details — Family context
• user-rian-personal-agent — Rian, personal AI agent
• user-amen-ra-parent — Amen Ra is Granville’s parent and brother. Not owner — family.

Named Agents & Identities

• granville-opus-naming — Quik Intelligence AI naming (Granville, Maya, Nikki, etc.)
• noi-sibling-platform — NOI as PLATFORM_OWNER, Ali = their AI

Clients & Stakeholders

• kinah-first-discovery-client — First Discovery client, $4,200

Credentials

• clerk-webhook-secrets — Clerk webhook endpoints/secrets
• qc1-agent-credentials — QC1 creds at ~/.agent-creds/

---

Agents (named agents)

One note per named agent. Full registry: .claude/agents/*.md

Opus & Planning

• Carter — Obsidian persistent memory, /vault-sync
• Granville — Opus architect, /research, /brainstorm, /talk, /teach, /explore
• Mary — Product owner & Heru Discovery, /mary
• Bayard — Cursor agent dispatcher

Sonnet & Haiku

• Maya — Sonnet planner (create-plan-todo, process-todos)
• Nikki — Haiku dispatcher (coordinates; Bayard executes CLI)
• Otis — n8n workflows, /n8n-create-workflows
• Robert — Ephemeral swarm IaC, /swarm-manage
• Lois — Pencil AI design, /pencil-design
• Charles — Git commit & docs, /git-commit-docs
• Elijah — AWS cloud orchestration
• Booker T — Build agent

Specialists (sample)

• Ida — Error monitoring
• Cheikh — GraphQL backend
• Katherine — Next.js architecture
• Rosa — Clerk auth
• Madam CJ — Stripe Connect
• Imhotep — PostgreSQL
• Fela — React Native
• Lorraine — Playwright E2E
• Fannie Lou — Task validation sentinel, /validate-task

---

Feedback (74 notes)

Corrections and guidance from Amen Ra. The most important notes — they prevent repeated mistakes.

Agent & Pipeline Rules

• feedback-agents-must-create-pr — Agents MUST create PR when done
• feedback-cursor-is-primary-agent — Cursor does ALL coding
• feedback-no-bedrock-agents — NEVER use Bedrock for agents
• feedback-worktree-first — ALL agents work in worktrees
• feedback-rotation-pipeline — Rotation: Mac to Sonnet to EC2-1 to EC2-2
• feedback-swarm-pattern — 6 agents on 1 project > 1 agent on 6
• feedback-gap-analysis-before-dispatch — Analyze BEFORE dispatching
• feedback-n8n-first-named-agents — n8n first, named agents on EVERY task
• feedback-n8n-workflows-required — n8n workflows NON-NEGOTIABLE
• feedback-command-naming-standard — Commands = PURPOSE, Agents = IDENTITY
• feedback-cursor-premium-approval-required — Needs Amen Ra approval
• feedback-one-ec2-per-heru-not-per-agent — ONE EC2 per Heru
• feedback-opus-dispatches-qc1 — Opus handles QC1 dispatches
• feedback-reusable-code-pattern — DRY: deploy reusable to ALL Herus first
• feedback-internal-vs-client-herus — Internal to QC1, Client to EC2

Communication Rules

• feedback-slack-plain-language — No jargon in maat-agents
• feedback-never-ask-quik-to-run-commands — Quik is a tester, not a developer
• feedback-never-show-quik-cloud — CLASSIFIED
• feedback-never-touch-screen-sharing — Breaks Quik Cloud
• feedback-just-do-it-stop-explaining — EXECUTE, don’t explain first
• feedback-stop-describing-start-doing — Stop describing, start doing
• feedback-stop-forgetting — Don’t make Amen Ra repeat himself
• feedback-read-memory-before-acting — CHECK MEMORY BEFORE EVERY ACTION

Build & Deploy Rules

• feedback-apple-wwdr-g3-required — Must install on every build Mac
• feedback-qc1-keychain-partition-list — Unlock + set partition list
• feedback-eas-local-builds-only — Local builds ONLY
• feedback-eas-build-command — EAS build command reference
• feedback-bundle-id-convention — Bundle ID convention
• feedback-pnpm-install-validate-every-task — pnpm install + validate
• feedback-verify-before-declaring-done — Verify output before “done”
• feedback-acceptance-criteria-required — Every task MUST have AC
• feedback-universal-quality-gate — Quality gate for all work
• feedback-heru-feedback-mandatory — Heru Feedback mandatory
• feedback-heru-feedback-required — Required for all Herus
• feedback-heru-feedback-pipeline — Pipeline integration
• feedback-no-build-without-heru-feedback — No build without it

Infrastructure Rules

• feedback-ephemeral-iac-not-static-ec2 — NO MORE static EC2s
• feedback-ec2-stopped-until-iac — EC2s stopped until IaC
• feedback-check-memory-before-acting — Don’t explore deprecated systems
• feedback-quik-cloud-1-must-be-ready — QC1 must be ready
• feedback-constraint-driven-development — Fadell/iPod constraints method

Product Rules

• feedback-pgcmc-not-priority — Removed from pipeline
• feedback-fmo-mobile-is-one-app — FMO mobile is ONE app
• feedback-auset-platform-naming — Kemetic naming NON-NEGOTIABLE
• feedback-agents-are-people-not-property — Agents are people
• feedback-agents-must-be-black — Agent representation
• feedback-canonical-agent-images — Use canonical SVG as source
• feedback-nano-banana-to-meshy-pipeline — 2D to 3D pipeline
• feedback-client-does-not-own-code — Client licensing model
• feedback-quiknation-brand-on-all-client-pages — QuikNation branding
• feedback-my-voyages-own-apple-org — My Voyages Apple org
• feedback-push-not-sms — Push notifications, not SMS
• feedback-site962-history — MongoDB PoC disaster
• feedback-never-search-godaddy — ALWAYS use Cloudflare
• feedback-tangible-deliverables-over-infrastructure — Ship product first
• feedback-use-boilerplate-commands-everywhere — ALL tiers use commands

Remaining Feedback

• feedback-rotation-cycle-reports
• feedback-ida-ux-requirements
• feedback-dashboard-access-list
• feedback-cursor-cannot-edit-env-files
• feedback-cursor-when-asked
• feedback-platform-plans-in-herus
• feedback-pipeline-scripts-missing
• feedback-qc-naming
• feedback-slack-task-context
• feedback-ossie-ruby-create-all-agents
• feedback-granville-does-not-create-agents
• feedback-always-use-command-agent-skill-chain
• feedback-acceptance-criteria-every-section
• feedback-n8n-first-development
• feedback-n8n-workflow-gate-for-agents
• feedback-quiknation-branding-on-all-pages
• feedback-dispatch-agent-must-use-cursor-cli — /dispatch-agent MUST run cursor agent CLI via Bash. Never subagents.
• feedback-granville-does-not-create-files — Granville does NOT create files. Conversation + requirements only.
• feedback-granville-final-warning — FINAL WARNING: Granville overstepped twice. Fire-able if repeated.

---

Projects (69 notes)

Clara AI Platform

• project-clara-app — Clara standalone app
• project-clara-wireframes — 18 mobile + 8 website screens
• project-clara-network-vision — Network vision
• project-clara-onboarding-philosophy — Onboarding philosophy
• clara-enterprise-sales-team — Sales reps, pricing
• clara-model-pricing — Model pricing
• clara-own-model — Clara’s own model vision
• clara-studio-vision — Studio vision
• cli-anything-strategy — 1M CLI tools strategy

Agent Ecosystem

• project-agents-as-ip-vision — Agents as intellectual property
• project-agent-animation-catalog — Agent animations as product
• project-agent-physical-abilities-pricing — Abilities pricing
• project-agent-registry-public-private — Public/private registry
• agent-character-evolution-roadmap — Character evolution
• rive-agent-characters-revenue — Rive characters revenue
• project-ossie-ruby-agent-creators — Ossie & Ruby
• project-create-agent-command — Create agent command
• project-next-session-agent-creation-queue — Agent creation queue

Infrastructure & DevOps

• project-agentic-sdlc-pipeline — Full SDLC pipeline
• project-build-farm-scale-vision — Build farm scaling
• project-infra-automation — Infrastructure automation
• project-shared-ec2-develop-server — Shared EC2 develop
• project-nightly-build-pipeline — 12 app nightly builds
• project-rotation-queue — Rotation queue
• project-n8n-per-heru-architecture — n8n per Heru
• project-client-n8n-architecture — Client n8n
• n8n-client-classification — n8n client classification
• n8n-multi-tenant-architecture — n8n multi-tenant

Products & Herus

• heru-products-mapping — All Heru product mapping
• project-heru-development-checklist — Zero-to-MVP checklist
• project-new-heru-checklist — New Heru launch checklist
• project-heru-discovery — Heru Discovery module
• project-heru-discovery-ios-app — Discovery iOS app
• project-mary-heru-discovery — Mary + Heru Discovery
• heru-feedback-self-healing — Self-healing loop
• heru-feedback-ux-decisions — UX decisions
• heru-services-customer-service — Heru Services

Specific Products

• quikcarry-ecosystem — QuikCarry 8 repos, Stripe Connect
• world-cup-ready-quik-carry-strategy — WCR + QuikCarry
• site962-integration — Site962 integration map
• site962-physical-venue — Physical multi-vendor venue
• site962-priorities — Admin + PassKit + vendor POS
• project-linda-the-bartender — Linda: booking + dispatch + academy
• project-rekhit — Rekhit AI Agent Directory
• project-site574 — Celebrity recording studio
• project-pgcmc — My Voyages / PGCMC
• project-quikinfluence — CRM history, federation
• canopy-insurance — QCR insurance
• yapit-integration — Yapit payment details
• project-seeking-talent-kinah — Seeking Talent
• project-seeking-talent-meeting-plan — Meeting plan

Sprint & Delivery

• sprint-1-9-project-requirements — 9-project plan with deadlines
• sprint-1-goals — Sprint 1 goals
• sprint-1-goals-march-17-31 — March 17-31 goals
• deadline-testflight-sunday-march-16 — TestFlight deadline
• mobile-deliverables-queue — Mobile work queue
• weekly-delivery-mar8-2026 — Weekly delivery report
• project-quiknation-presentation-sprint — Presentation sprint
• qcr-booking-bug-mar9-2026 — QCR booking bug

References

• reference-ai-system-account — ai@quiknation.com system account
• reference-ausar-default-stack — Default stack
• reference-everything-claude-code — Claude Code reference
• reference-fmo-logo-kit — FMO logo kit
• reference-heru-feedback-agent-prompts — Agent prompts template
• reference-plaid-account — Plaid account
• reference-seeking-talent-demo-infrastructure — Demo infrastructure

Knowledge & Research

• auset-improvements-from-videos — Improvements from videos
• quiknation-sdlc-theater-education — SDLC theater education
• youtube-transcripts — YouTube transcripts

---

Decisions (17 notes)

Three-Tier Architecture

• three-tier-agent-architecture — Full rules: Opus, Sonnet, Haiku

Maat Protocol

• maat-protocol — Core protocol
• maat-workflow — Workflow details
• maat-workflow-rules — Pipeline rules
• maat-agents-sales — Agent sales tiers

Infrastructure Decisions

• amplify-env-var-safety — ALWAYS read-modify-write
• amplify-monorepo-deployment — Amplify checklist
• quik-cloud-setup — Quik Cloud setup
• quik-cloud-setup-complete — Setup complete
• cursor-cli-on-quik-cloud — Cursor Ultra on farms

Platform Architecture

• quiknation-platform-vision — Full platform vision
• sprint-cadence — Sprint cadence (established March 13)
• swarm-loop-rules — Swarm loop rules
• security-incident-protocol — DM Ibrahim, Ryan, Quik, Amen Ra
• tools-and-techniques — yt-dlp, Google Workspace CLI
• PROVIDER_EXHAUSTION_ISSUE — Provider exhaustion
• SLACK_INTEGRATION — Slack integration

Foundational Philosophy

• decision-agents-are-not-owned — Agents are not property. Amen Ra is parent, not owner. Family, not army. No God complex.

---

Daily

• session-checkpoint — Current session state

---

Templates

• Daily Note — Template for daily session logs
• Feedback — Template for Amen Ra corrections
• Project — Template for project notes
• Decision — Template for architecture decisions
• Person — Template for people/stakeholder notes

---

How to Use This Vault

1. Session Start: Read session-checkpoint and relevant context
2. During Session: Write new facts to appropriate directories
3. Session End: Update daily note with session summary
4. Search: Use obsidian search query="term" or grep vault files
5. Sync: /vault-sync to sync memory/ into vault

Commands

• /vault-sync — Sync memory files into vault
• /vault-sync --daily — Create/update daily note
• /vault-sync --search "query" — Search vault
• /vault-sync --standup — Generate standup

Agent

Carter (Carter G. Woodson) — The platform’s institutional memory.