Customizing UCIS Dimensions
Relevant source files
The following files were used as context for generating this wiki page:
The Ultimate Content Intelligence & Implementation System (UCIS) is a framework designed to deconstruct content into structured intelligence products. In its v5.1 "Monetization & Commercial Yield Edition," the system utilizes 11 distinct Knowledge Dimensions to parse transcripts and metadata into actionable data for specific target personas.
Customizing these dimensions involves configuring the persona weights, cognitive lenses, and the underlying prompt structure stored within the system's database. This allows for tailored extraction strategies ranging from executive summaries (Apex Intelligence) to deep commercial viability assessments (Monetization Profiling).
Sources: web/lib/prompts/ucis-v5.1.ts:1-30, supabase/migrations/20260611142500_add_prompt_config.sql:4-20
UCIS Framework Architecture
The UCIS system operates on a "Closed Universe" principle, where the provided transcript is the absolute source of truth. The architecture transitions from raw input to a structured JSON output through a series of internal processing steps, including metadata ingestion, insight ranking, and persona-based depth allocation.
Processing Pipeline
The following diagram illustrates the flow from content ingestion to the final 11-dimension structured output.
The 11 Knowledge Dimensions
The system is hard-coded to expect 11 dimension headers. If a transcript lacks sufficient data for a specific dimension, the system must invoke the Insufficient Data Protocol rather than inventing content.
| Dimension | Name | Primary Focus | Target Persona |
|---|---|---|---|
| 1 | Apex Intelligence | 60-second ROI & Executive Summary | All (Weighted) |
| 2 | Provenance & Metadata | Authority markers & Engagement metrics | Content Creator |
| 3 | Content Architecture | First Principles deconstruction | Consultant/Researcher |
| 4 | Psychological Layer | Sentiment, Tonal Profile, & Bias Detection | Analyst |
| 5 | Core Intelligence | Tier 1 Breakthrough Insights & Quotes | All |
| 6 | Comparative Analysis | Performance metrics in tabular format | All |
| 7 | Implementation Systems | Executable step-by-step workflows | Indie Maker |
| 8 | Semantic Foundation | Knowledge Graph nodes and edges | RAG Systems |
| 9 | Strategic Foresight | Trend projections & Unfair advantages | All |
| 10 | Credibility & Risk | Meta-assessment & Disclosures | All |
| 11 | Commercial Yield | Monetization & Revenue Potential | Creator/Indie Maker |
Sources: web/lib/prompts/ucis-v5.1.ts:69-325, web/lib/utils/ucis-parser.ts:4-18, scripts/verify-parser-v5.1.ts:32-45
Persona Configuration & Weighting
A critical aspect of customizing UCIS is the Persona Header. This configuration dictates the depth of analysis allocated to each dimension. The primary persona is assigned a 50% weight, affecting the utility functions of the extracted deliverables.
Weighted Distribution
The system supports five ranked target personas:
- P1 Content Creator: Focused on RPM, audience growth, and virality.
- P2 Indie Maker: Focused on lead generation and affiliate scalability.
- P3 Consultant: Focused on advisory retainers and service positioning.
- P4 Researcher: Focused on academic/grant positioning.
- P5 Product Manager: Focused on internal alignment and thought leadership.
Sources: web/lib/prompts/ucis-v5.1.ts:55-67, web/lib/prompts/ucis-v5.1.ts:275-300, web/lib/tests/ucis-v5-validator.test.ts:37-55
Implementation & Parsing Logic
The system uses a robust index-slicing approach to parse raw markdown into structured dimension objects. The parseToUCISDimensions function identifies headers using a specific regex pattern and extracts content between them.
// web/lib/utils/ucis-parser.ts:25-45
export function parseToUCISDimensions(
markdown: string | null | undefined,
): Record<number, UCISDimension> {
const headerRegex = /^###\s+DIMENSION\s+(\d+)\b[^\n]*/gim;
// ... sorts and slices content based on detected header indices
}
Sources: web/lib/utils/ucis-parser.ts:25-63
Validation Protocol
Every UCIS output is subjected to a validation check (implemented in UCISValidator) to ensure structural integrity:
- Structural Check: Verifies the presence of the Persona Header and all 11 dimension headers.
- Rigor Check: Counts in-content timestamps (formatted as
HH:MM:SS) to ensure quantitative claims are sourced. - Emoji Restriction: Forbids all emojis except the warning symbol (⚠️) in risk disclosure blocks.
Sources: web/lib/tests/ucis-v5-validator.test.ts:5-100
Database Configuration
The prompt configuration is persisted in the app_settings table via Supabase migrations. This allows the system to maintain a versioned history of UCIS prompts, such as the transition from v5.0 (Persona-Weighted) to v5.1 (Monetization Edition).
-- supabase/migrations/20260611142500_add_prompt_config.sql:1-10
insert into public.app_settings (key, value)
values (
'prompt_config',
'{"latest":"5.1","history":[{"version":"5.1", ...}],"versions":{"5.1":"..."}}'
);
Sources: supabase/migrations/20260611142500_add_prompt_config.sql:1-15
Conclusion
Customizing UCIS Dimensions provides a flexible yet rigid framework for content intelligence. By modifying the persona weights in the header and the scoring algorithms in the system prompt, developers can redirect the focus of the "ruthless parser" to serve different commercial or analytical objectives while maintaining the technical constraints of the "Closed Universe" logic.
Sources: web/lib/prompts/ucis-v5.1.ts:36-53, web/lib/prompts/ucis-v5.1.ts:357-366