yogsoth-ai Dimensional Analysis
Map the design space of a research area by discovering its fundamental dimensions (axes of variation), enumerating meaningful combinations, and identifying unexplored regions that represent novel opportunities.
Manifest
| Level | Count | Skills |
|---|---|---|
| Strategy | 3 | dimension-discovery, combination-mapping, gap-prioritization |
| Tactic | 2 | axis-extraction, matrix-generation |
| SOP | 6 | dimension-page-creation, axis-validation, combination-enumeration, novelty-scoring, question-generation, matrix-export |
Budget Table
| Metric | Small | Medium | Large |
|---|---|---|---|
| Dimensions identified | 3 | 6 | 10 |
| Axes per dimension | 3 | 5 | 8 |
| Combinations explored | 10 | 30 | 60 |
| Novel gaps identified | 3 | 8 | 15 |
| Questions generated | 5 | 12 | 25 |
Strategy Sequence (Reference, Not Prescription)
- dimension-discovery — identify fundamental axes of variation in the design space
- combination-mapping — enumerate interesting combinations, mark existing work
- gap-prioritization — rank unexplored combinations by novelty and feasibility
MCP Tools Used
vault_search— find existing coverage of dimension valuesvault_add_edge— connect dimensions to conceptsvault_query_graph— explore existing dimension neighborhoodsvault_graph_stats— assess coverage completeness
Context-Management
Guiding Principles
- Dimensions are independent. If two axes always co-vary, they're likely one dimension.
- Completeness over depth. Map the full space first, then drill into promising regions.
- Empty cells are opportunities. An unexplored combination in a well-populated matrix is a research opportunity.
- Feasibility constrains. Not all combinations are realizable. Mark impossible combinations explicitly.
- Cross-domain dimensions. The most interesting dimensions often come from outside the domain.