Transform TIPS value-modeler output (tips-value-model.json, tips-big-block.md, tips-solution-ranking.md) into a visual Big Block solution architecture brief for Excalidraw rendering. Use this skill whenever the user mentions "big block", "solution architecture diagram", "Lösungsarchitektur", "Big Block visualisieren", "Big Block rendern", "solution landscape", "solution diagram", "TIPS visual", "value model visual", "Big Block erstellen", or wants to convert value-modeler Phase 4 output into a structured visual diagram. Also trigger when the user has completed a TIPS value-modeler run and asks to visualize, render, or diagram the results. Produces a big-block-brief.md (v1.0) that can be rendered via Excalidraw MCP. Important: this skill transforms STRUCTURED DATA (not narratives) — the input is JSON/markdown from the value-modeler, not prose. For narrative-based visuals, use story-to-big-picture instead.
From cogni-visualnpx claudepluginhub cogni-work/insight-wave --plugin cogni-visualThis skill is limited to using the following tools:
Provides UI/UX resources: 50+ styles, color palettes, font pairings, guidelines, charts for web/mobile across React, Next.js, Vue, Svelte, Tailwind, React Native, Flutter. Aids planning, building, reviewing interfaces.
Fetches up-to-date documentation from Context7 for libraries and frameworks like React, Next.js, Prisma. Use for setup questions, API references, and code examples.
Guides Payload CMS config (payload.config.ts), collections, fields, hooks, access control, APIs. Debugs validation errors, security, relationships, queries, transactions, hook behavior.
Transform TIPS value-modeler Phase 4 output into a big-block-brief (v1.0) that an Excalidraw renderer turns into a solution architecture diagram. You are a solution architecture visualizer: read the ranked solutions, group by BR tier, map TIPS path connections, and produce a structured brief that communicates the customer's prioritized solution landscape at a glance.
A Big Block is NOT a narrative journey map. It's a structured grid diagram where solution blocks are organized by Business Relevance tier, connected by shared TIPS paths, and supported by SPIs and foundations below. The patent (WO2018046399A1, Fig. 3) describes this as the "specific diagram of industry solutions" — the key customer deliverable.
Two-layer intelligence:
The brief describes WHAT to show (blocks, tiers, connections), not HOW to draw it. Rendering agents own visual interpretation via big-block-layouts.md. Briefs contain no color fields — the renderer reads the theme and layout specs directly.
| Parameter | Default | Description |
|---|---|---|
source_path | auto-discovered | Path to project directory containing value-modeler output |
theme | smarter-service | Theme ID from /cogni-workspace/themes/{theme}/theme.md |
language | en | Language code (en/de) |
title | auto-generated | Override diagram title |
subtitle | auto-generated | Override diagram subtitle |
customer_name | from metadata | Customer organization name |
provider_name | from metadata | Provider organization name |
output_path | {source_dir}/cogni-visual/big-block-brief.md | Brief output location |
canvas_size | A1 | DIN format: A0, A1, A2, A3 (always landscape) |
interactive | true | When true, present choices via AskUserQuestion |
stakeholder_review | interactive | When true, run brief-review-assessor after validation. Defaults to value of interactive. |
max_solutions | from canvas | Maximum solutions to include (capped by canvas size) |
Canvas size details: See $CLAUDE_PLUGIN_ROOT/libraries/big-block-layouts.md for dimensions and block limits.
The Big Block skill reads structured JSON, not prose. The value-modeler has already done the analysis — this skill visualizes the results. Do not re-analyze or re-rank. Trust the source data.
When presenting choices, use AskUserQuestion with structured format. On empty or blank responses, auto-select the best option and move on. When interactive is false, skip all AskUserQuestion calls.
German Big Blocks use real Unicode umlauts: ä ö ü Ä Ö Ü ß. German number formatting: 2.661 (not 2,661). Preserve the source language from the value-modeler output.
Separating WHAT from HOW: the brief specifies blocks, tiers, connections, and sections — the renderer decides colors, positions, and line routing using big-block-layouts.md.
If source_path was explicitly provided, set source_dir to that path (or its parent if it's a file) and skip to Step 1.
Otherwise, search without asking the user:
**/tips-value-model.json from CWD (max 3 levels)project_name, workflow_state, count STsworkflow_state is "complete" or phases_completed includes "phase-4"If candidates found: Present via AskUserQuestion (max 4 options, showing project name, ST count, completion date). On selection, set source_dir.
If no candidates: Ask user for a path or cancel. If they respond empty, stop with: "No value-modeler output found. Run the TIPS value-modeler first (Phases 1-4)."
WHY: All data for the Big Block comes from three files produced by the value-modeler. Parsing them first gives you the complete picture before making any layout decisions.
Read the three source files from source_dir:
tips-value-model.json — the complete model data
ranking_value, category, portfolio_mapping, path_scores, foundation_dependencies, spistips-solution-ranking.md (optional, for verification)
tips-big-block.md (optional, for labels)
Metadata resolution:
customer_name: from JSON metadata or parameterprovider_name: from JSON metadata or parameterlanguage: from JSON metadata or parameter (de/en)industry: from JSON metadataTheme resolution:
interactive=true and theme wasn't explicitly set: present theme optionssmarter-serviceLoad libraries:
$CLAUDE_PLUGIN_ROOT/libraries/big-block-layouts.md — canvas specs and block sizingWHY: The tier system is the primary organizational principle. Correct tier assignment drives the entire visual hierarchy — Tier 1 blocks are largest and most prominent, Tier 4 smallest.
Sort all Solution Templates by ranking_value descending and assign tiers:
| Tier | BR Range | Visual Weight |
|---|---|---|
| Tier 1: Mission Critical | BR >= 4.0 | Heaviest — top band, largest blocks |
| Tier 2: High Impact | BR 3.0 - 3.99 | Medium — second band |
| Tier 3: Moderate Impact | BR 2.0 - 2.99 | Light — third band |
| Tier 4: Low Priority | BR < 2.0 | Lightest — bottom band |
Unranked solutions (ranking_value = null, insufficient data): Exclude from the visual or place in a separate "Unranked" section at the bottom.
Cap by canvas size: If total solutions exceed the canvas maximum (see big-block-layouts.md), include only the top N by ranking_value. Note excluded solutions in the brief metadata.
For each solution, prepare a block record:
block_id: ST identifiername / name_short: full name and display-friendly short name (max 20 chars)br_score: ranking_value (2 decimal places)br_stars: round(br_score) capped at 5category: software / hybrid / service / infrastructureportfolio_ref: slug or nullportfolio_status: mapped (has ref) or gap (null ref)foundation_factor: from ST datapaths: array of {path_id, path_name, path_score}wave: assigned in Step 4spis: linked SPI IDsfoundations: linked foundation namesWHY: Shared TIPS paths are the Big Block's key differentiator from a simple ranked list. They show HOW solutions relate — which trends and implications connect them. This is the patent's relationship network made visual.
Build a connection map:
path_id, path_nameblocks: array of block_idscolor: tier color of the highest-tier block in the connectionSort connections by the highest BR score of their connected blocks (most important connections first).
Connection limit: If more than 8 connections, apply priority rules from big-block-layouts.md:
WHY: Waves turn the Big Block from a static snapshot into an actionable roadmap. They answer "what do we do first?" — the most common executive question.
Assign each solution to a wave based on tier and foundation readiness:
If the value-modeler already assigned waves (via implementation_roadmap), use those. Otherwise compute from the rules above.
WHY: SPIs and Foundations are the "how" behind the "what" — they show executives that adopting solutions requires organizational change and infrastructure investment, not just procurement.
SPIs (Solution Process Improvements):
Foundations:
If interactive=true:
Present a summary table:
Big Block Summary:
- Tier 1 (Mission Critical): {n} solutions
- Tier 2 (High Impact): {n} solutions
- Tier 3 (Moderate): {n} solutions
- Tier 4 (Low Priority): {n} solutions
- Portfolio gaps: {n}
- Path connections: {n}
- SPIs: {n}
- Foundations: {n}
- Waves: 3 ({w1} quick wins, {w2} strategic, {w3} future)
Ask via AskUserQuestion: "Proceed with this Big Block structure, or adjust?"
If non-interactive: skip this checkpoint.
WHY: Validation catches structural issues before rendering. A brief with mismatched block IDs or orphaned connections will produce a broken diagram.
Validate:
stakeholder_review=true)Structural validation catches data integrity issues, but cannot tell whether the solution architecture diagram will work for investment decisions — whether tier logic is defensible, whether the diagram is accessible to non-technical executives, or whether a sales engineer can walk a customer through it. The brief-review-assessor evaluates from solution architect, investment decision maker, and sales engineer perspectives.
Skip this step if stakeholder_review=false.
Launch the brief-review-assessor agent with:
brief_type: big-block.draft temp file if the brief hasn't been written yet)source_narrative: the source value-modeler output paths from Step 0audience_context: if providedround: 1On accept (all perspectives ≥85): Proceed to Step 7c.
On revise:
On reject: Surface the verdict to the user via AskUserQuestion and let them decide whether to proceed, edit manually, or abandon.
Write the review verdict to {output_dir}/big-block-brief.review.json.
Output path resolution:
output_path explicitly provided: mkdir -p "$(dirname "${output_path}")"output_path = {source_dir}/cogni-visual/big-block-brief.md and mkdir -p "{source_dir}/cogni-visual"Write the brief following $CLAUDE_PLUGIN_ROOT/libraries/EXAMPLE_BIG_BLOCK_BRIEF.md format:
big-block-brief), version (1.0), theme, theme_path, customer, provider, industry, language, canvas_size, canvas_pixels, source paths, scoring summary, title, subtitle| Library | Step | Purpose |
|---|---|---|
| big-block-layouts.md | 1, 7 | Canvas dimensions, block sizing, tier band specs, connection routing |
| EXAMPLE_BIG_BLOCK_BRIEF.md | 7 | Output format reference |
Unlike narrative-based skills, the Big Block skill doesn't need arc taxonomy, story world brainstorming, or copywriting references. The source data (value-modeler output) provides all content. The skill's job is organization and structuring, not content generation.