Content OS

Make the control dashboard website responsive so that it can be easily used on smartphones

1.11.4 Self-Hosted LLaMA 3.3 70B Gateway & Groq Failover

Integrate NVIDIA DGX Spark self-hosted LLaMA 3.3 70B NVFP4 as primary LLM engine with automatic failover to Groq API backup and heuristic circuit-breaker. Acceptance test: npm run test:llm.

1.11.3 Automated Server & UI Integration Verification Suite

Author automated HTTP integration test suite validating all REST endpoints, sprint triggering, and approval decisions. Acceptance test: npm run test:server.

1.11.2 Modern Fluid Motion Web UI & Interactive Control Plane

Build interactive single-page dashboard with glassmorphism, fluid spring animations, Chief of Staff studio, DAG visualizer, approval queue, and multi-tenant switcher. Acceptance test: npm run test:server.

1.11.1 Command Center HTTP REST API & Event Stream Server

Implement REST API server exposing endpoints for status, creators, workflows, approvals, chat, sprints, and graphs. Acceptance test: npm run test:server.

1.10.4 Stage-Gate 5 Project Closeout & APMF Lessons Learned Postmortem

Execute Stage-Gate 5 Closeout Review in Operion, generate closeout report, and compile LESSONS_LEARNED_POSTMORTEM.md.

1.10.3 Second Persona Onboarding & Cross-Tenant Non-Contamination Verification Suite

Define Dr. Elena Vance persona, author comprehensive multi-tenant test suite proving zero data leakage and domain-agnostic execution. Acceptance test: npm run test:productize.

1.10.2 Creator Instance Onboarding Engine & Multi-Tenant Registry

Implement CreatorOnboardingEngine to provision tenant directories, databases, and register instances. Acceptance test: npm run test:productize.

1.10.1 Core vs Instance Architectural Separation & Tenant Isolation

Implement CreatorInstance encapsulating isolated storage, graphs, and Chief of Staff orchestrators per tenant. Acceptance test: npm run test:productize.

1.9.7 Automated Chief of Staff Verification Suite & Interactive CLI Terminal

Create comprehensive test suite validating all Chief of Staff intents, approvals, and build interactive CLI REPL. Acceptance test: npm run test:orchestrator.

1.9.6 Chief of Staff Conversational Orchestrator & Command Center

Implement ChiefOfStaffOrchestrator coordinating workflows across departments, enforcing governance and approval gates, and logging runs. Acceptance test: npm run test:orchestrator.

1.9.5 Chief of Staff Intent Parser & Routing Engine

Implement natural language intent classification and parameter extraction routing creator requests across departments. Acceptance test: npm run test:orchestrator.

1.9.4 Automated Learning Loop Verification Suite & CLI Runner

Create comprehensive test suite validating learning loop, event ingestion, and CLI digest runner under 0/100 EVM criteria. Acceptance test: npm run test:learning.

1.9.3 Weekly Strategy Digest & \"What Worked\" Reporter

Build weekly executive digest generator synthesizing winning component recipes, fatigue rotations, and strategy directives. Acceptance test: npm run test:learning.

1.9.2 Closed-Loop Learning Engine & Hypothesis Validator

Implement closed-loop learning cycle (Observation -> Candidate -> Validation -> Operationalization) updating component tiers and strategy weights. Acceptance test: npm run test:learning.

1.9.1 Performance Event Ingestion & Component Attribution Tracker

Ingest performance metrics across platforms, map retention curves to atomic components, and record performance events. Acceptance test: npm run test:learning.

1.8.4 Automated Omnichannel Verification Suite

Create comprehensive test suite validating all platform adapters, constraints, and multiplication pipeline under 0/100 EVM criteria. Acceptance test: npm run test:omnichannel.

1.8.3 Omnichannel Multiplication Pipeline & Campaign Generator

Build OmnichannelMultiplier coordinating many-to-many reconstruction across all target platforms with provenance preservation. Acceptance test: npm run test:omnichannel.

1.8.2 Format Architects & Platform Adapters

Implement native format adapters transforming semantic DAG nodes into LinkedIn Posts/Carousels, X Threads, Newsletters, and Instagram Reels. Acceptance test: npm run test:omnichannel.

1.8.1 Platform & Format DNA Knowledge Engine

Build Platform DNA engine defining constraints, attention patterns, structures, and CTA conventions for YouTube Shorts, LinkedIn, X, Instagram Reels, and Newsletters. Acceptance test: npm run test:omnichannel.

1.7.4 Navigable Provenance Engine

Build full backward and forward provenance graph traversal (Claim -> Evidence -> Source -> Component -> Script). Acceptance test: npm run test:graphs.

1.7.3 Performance Graph & Second-Level Component Attribution

Build Performance Graph with granular retention attribution (3s, 10s, 25%, 50%, completion) mapped to specific atomic components. Acceptance test: npm run test:graphs.

1.7.2 Content Intelligence Graph & Semantic DAG Architecture

Build Content Intelligence Graph (Idea, Claim, Evidence, Hook, Component, Asset) with REGENT-style recursive edge operations (CREATE, EXTEND, QUALIFY, CONTRADICT, LINK). Acceptance test: npm run test:graphs.

1.7.1 Groq LLM Client & Evidence Extraction Engine

Build resilient Groq LLM client (gpt-oss-120b / qwen3.8-27b) with structured JSON extraction for claims, evidence, and component tags. Acceptance test: npm run test:graphs.

1.6.3 Workflow Runtime & Vertical Slice Runner CLI

Implement workflow state machine (draft, running, waiting_approval, completed) and CLI runner. Acceptance test: npm run test:pipeline.

1.6.2 Script Production Agent & Psychology Checklist

Build Script Team agent with component breakdown and psychology checklist QA validation. Acceptance test: npm run test:agents.

1.6.1 Opportunity Intelligence & Strategy Agents

Build Opportunity Intelligence and Editorial Strategy agents with structured Zod outputs. Acceptance test: npm run test:agents.

1.5.1 Psychology Engine & Behavioral Funnel Library

Implement behavioral theory rules (Curiosity Gap, Zeigarnik Loops, Cognitive Fluency) and QA scoring. Acceptance test: npm run test:psychology.

1.4.3 Recombination Matrix & Recipe Engine

Implement algorithmic recombination of components into winning recipes based on performance memory. Acceptance test: npm run test:components.

1.4.2 Video & Content Decomposer Engine

Implement video, transcript, and competitor content parser into constituent component buckets. Acceptance test: npm run test:components.

1.4.1 Component Schema & Bucket Store

Implement atomic component schemas (Hook, Topic, Narrative, Visual, Audio, Presentation) and catalog buckets. Acceptance test: npm run test:components.

1.3.2 Creator Graph Query Engine

Implement Creator Graph query engine with bounded context extraction for agents. Acceptance test: npm run test:creator.

1.3.1 Creator Constitution Schema & Aryan's Profile

Author structured Creator Constitution for Aryan / Revedor (identity, voice, beliefs, pillars, boundaries). Acceptance test: npm run test:creator.

1.2.2 SQLite Database Schema & DAO Layer

Build local-first SQLite database schema with Drizzle/better-sqlite3 and DAO layer. Acceptance test: npm run test:db.

1.2.1 Core Ontology & TypeScript Type Definitions

Define canonical domain types and Zod schemas for the 8 entity families. Acceptance test: npm run test:schema.

1.1.2 Operion Workstream & Stage-Gate Synchronization

Sync 9 WBS Workstreams and map Stage-Gate milestones into Operion via MCP.

1.1.1 APMF Governance & Skills Setup

Install APMF skills into .gemini/skills and set governance rules in .gemini/rules/project_management.md.

Split Core vs. Creator Instance

Separate department/agent logic (Core) from Creator Constitution + graph data (Instance); test the split by onboarding a second creator persona.

Build the orchestrator agent

Single conversational entry point that decomposes a creator request and routes to Research/Strategy/Creation/Psychology/Omnichannel/Distribution/Analytics agents.

Close the feedback loop

Component-level performance tagging → Performance Graph → weekly digest fed back into Editorial Strategy. This is what makes the org self-improving.

Build platform adapter agents

One spec-driven adapter agent per target platform, built and git-persisted session-by-session per the Canopy/long-horizon-agent skill.

Point REGENT at Content OS data

Extend REGENT's schema to ingest content components (hook, topic, format, performance) instead of standing up a second graph engine.

Build the 3-agent vertical slice

Opportunity Intelligence agent → Editorial Strategy agent → Script Team agent, chained end to end, with a hard-coded psychology checklist as a gate before output. No orchestrator yet — just prove the chain works for one platform.

Draft Creator Constitution spec

Write Identity, Domain, Voice, Content Philosophy, Audience, and Goals as a structured (not prose) spec — the schema every later department/agent prompt reads from.