Ecosystem Architecture

Totem Protocol

Open-source framework for AI-assisted structural intelligence. Created at Sense Collective. Licensed by Shur Creative Partners. The shared foundation both organizations build on.

Sense Collective

Think-tank research studio. Co-founded by Jonny Dubowsky, David Muller, and Itaal Shur (investor). Focus: interaction design, installation art (Perceptagon), educational programming, workshops, R&D at the edge. Raises money and pursues grants.

Shur Creative Partners

Licenses Totem Protocol and builds novel commercial capabilities on top. Agency delivering structural intelligence to clients through editorial briefings, visualization hubs, and brand power scoring.

SHUR IQ

First fully functioning product (from Shur Creative Partners). AI-powered structural intelligence engine. Diagnoses how brands, organizations, and industries work.

ShurAI

The agency brand and service offering. Client-facing identity for Shur Creative Partners' commercial work.

11
Active Projects
33
Deployed Skills
22
Mapped Capabilities
150
Entities
738
Relationships
15
Clusters
13
Cloudflare Sites

Client Portfolio

ClientStatusIntakeBrand PowerTop OpportunityDeliverable
AHA Integration 3.39/5.0 Friction analysis Editorial Brief
FrameBright Intelligence 14/20 Good Housekeeping Seal Editorial Brief
Fiserv Intelligence 11/20 46/100 Prosperity narrative Editorial Brief
AFDVI Intelligence 2.24/5.0 The Invisible Impact Intelligence Brief
MicroCo Complete Selected: aTwist
CondoSales Active POC 17 Social Intelligence
Champ School POC 16 Digital Twin
Habit Finder Discovery 18 Client decision

Commercial Model

1

Structural Signal Intelligence

$40K – $75K · 3–5 weeks

Intake scoring, gap analysis, brand power score, editorial brief, visualization hub.

2

Structural Deep Dive

$90K – $150K · 6–10 weeks

Multi-viewport intelligence, competitive stack ranking, content factory pipeline, strategic recommendations.

3

Architecture Blueprint

$125K – $250K · 8–12 weeks

Full organizational intelligence architecture, agent deployment, custom dashboards, integration plan.

4

Signal Monitoring

$8K – $20K/mo · 12-month minimum

Ongoing structural monitoring, gap detection alerts, quarterly deep reviews, evolving intelligence layer.

Year 1 target: $1.92M

Capability Map

22 capabilities decomposed from the Totem Agent Framework, organized by functional category.

Data Collection 3

Message Archival
Data Collection Agent
message-logger
Active
Link Processing
Data Collection Agent
link-processor
Active
Trend Monitoring
Gap Minder Agent
trend-monitor
Active

Intelligence 7

Content Classification
Data Collection Agent
classifier
Active
Semantic Search
Intelligence Agent
semantic-search
Active
Action Extraction
Intelligence Agent
action-extractor
Active
Insight Synthesis
Intelligence Agent
insight-synthesizer
Active
AI Analysis
Intelligence Agent
ai-analysis
Active
Knowledge Graph
Knowledge Agent
knowledge-graph
Eval 80%
Ontology Management
Knowledge Agent
ontology-manager
Eval 96%

Integration 1

Knowledge Export
Knowledge Agent
knowledge-export
Active

Content Creation 4

Content Factory
Content Factory Agent
content-factory
Active
Anti-Slop Content Generation
Content Factory Agent
anti-slop-pipeline
Active
Brand Naming Process
Content Factory Agent
brand-naming
Active
GPT Generation
Content Factory Agent
gpt-gen
Active

Gap Analysis 3

Scout Agent / Gap Analysis
Scout Agent
scout-gap-analysis
Active
Negative Space Exploration
Scout Agent
negative-space
Active
Gap Detection
Gap Minder Agent
gap-detection
Active

Orchestration 3

Agent Handoff Protocol
Orchestrator Agent
agent-handoff
Active
Client Intake-to-Delivery
Orchestrator Agent
intake-pipeline
Active
Intent Protocol
Orchestrator Agent
intent-protocol
Active

Memory 1

Dynamic Knowledge Repository
Memory Agent
dynamic-knowledge-repo
Active

Knowledge Graph

Force-directed graph of the Totem Protocol ecosystem. 150 nodes, 738 relationships, 15 clusters. Betweenness-scaled node size reveals structural bridges.

Structural Gaps

Intelligence Insights ↔ Methodology Metrics

Competitive intelligence outputs (editorial longform, PDFs, dashboards) lack structural connections to scoring methodology (brand power score, pentagon, radar, revenue model). Reports arrive without exposing how conclusions were reached.

Intelligence Insights ↔ Framework Research

Intelligence products (competitive analysis, surge reports, deployed sites) are disconnected from the underlying framework research (visualization research, totem framework, parallel agents). The delivery layer doesn't feed back into R&D.

Space Dynamics ↔ Communication Tools

Core analytical work (ShurAI, Palantir positioning, negative space, content factory) doesn't flow cleanly to distribution tools (Slack, Google Docs, handoff documents). Synthesis happens in one cluster, sharing happens in another.

Gap Analysis

Recurring structural gaps across sessions, confirmed independently through InfraNodus graph analysis and textual synthesis.

Gap 1: Intelligence Delivery ↔ Methodology

Intelligence reports (editorial longform, PDFs, dashboards) are disconnected from the scoring methodology (brand power score, pentagon, radar). The delivery format doesn't expose how conclusions were reached. Clients receive outputs without the reasoning chain.

Gap 2: Space Dynamics ↔ Communication Tools

The core analytical work (ShurAI, Palantir positioning, negative space, content factory) doesn't flow cleanly to distribution tools (Slack, Google Docs, handoff documents). Synthesis happens in one cluster, sharing happens in another. Every handoff is manual.

Gap 3: Narrative Strategy ↔ Ecosystem Architecture

Combined-v1 synthesis rules (honest framing, anti-AI voice) exist in isolation from the underlying capability framework. The editorial voice isn't architecturally connected to the capability map. Narrative decisions don't propagate to agent behavior.

Gap 4: Observer Layer Isolation

The content-map-updater produces gaps but those gaps don't feed back into synthesis. The observer layer is structurally isolated from the work it observes. Gap detection runs, generates findings, and files them. Nothing consumes the output programmatically.

Pattern

All four gaps share a structural signature: production and distribution are decoupled. Intelligence gets produced but doesn't flow. Methodology exists but doesn't surface. Observations accumulate but don't feed back. The system generates value at each node but the edges between nodes are weak or missing.

Revenue Pipeline

Conversion Funnel

Tier 1 → Tier 2
65%
Tier 2 → Tier 3
75%
Tier 3 → Monitoring
75%

Year 1 Revenue Targets

TierAvg PriceVolumeRevenue
Tier 1: Signal Intelligence$60K6$360K
Tier 2: Deep Dive$120K4$480K
Tier 3: Architecture Blueprint$180K3$540K
Layer 4: Signal Monitoring$180K/yr3$540K
Total$1.92M

Conservative scenario (5/3/2/2 split): $1.38M

SBPI Scoring Model

DimensionWeight
Content Strength20
Narrative Ownership20
Distribution Power25
Community Strength20
Monetization Infrastructure15
85–100 Category Dominant 70–84 Strong Ecosystem 55–69 Emerging Power 40–54 Niche <40 Limited

Session Intelligence

DateTimeProjectSummary
2026-03-1401:00shur, fiservBusiness Model Framework v2 synthesis + Fiserv editorial sanitization
2026-03-1423:30fiserv, bi-report-labMCP backlog complete, combined-v1 BI Report Lab, Fiserv editorial rewrite
2026-03-1313:00shurBI Report Lab — 10 parallel agents, 5 feedback branches, comparison sites
2026-03-1023:30totem-orchestrationMCP config fix (enabledMcpjsonServers allowlist)
2026-03-1015:00framebrightFeedback rounds, agent decision log, Meadows leverage points
2026-03-1014:00framebrightHoward Postley feedback, CRA business model, dynamic access control
2026-03-0921:00shurGWS CLI migration, 4 new skills, Gmail API enabled
2026-03-0914:00shurBizDev IQ call analysis, REA semantics, engagement cycle tracker
2026-03-0600:30totem-frameworkViz research consolidation, vault reorganization
2026-03-0514:00shurCloudflare migration (13 sites), deployed sites registry

Key Learnings

"InfraNodus graph gaps independently confirm textual analysis."
2026-03-14
"Anti-slop enforcement must apply to ALL generated text — Slack messages, handoffs, commit messages, everything."
2026-03-14
"Merging 5 experiment branches into combined-v1 produces a reusable standard."
2026-03-14
"Political sanitization: named individuals become role-based references."
2026-03-14