The Spectral Brand Audit Toolkit
Six structured prompts that turn any large language model into an SBT analyst. Each module accepts a brand name and produces a standardized YAML report. The full pipeline compresses a multi-week consulting engagement into a single analytical session.
How It Works
Copy a prompt module from GitHub. Paste it into any capable LLM (Claude, GPT-4, Gemini). Specify your brand. The model produces structured YAML output matching the corresponding template. Run all six modules sequentially for a complete spectral audit.
The pipeline has been validated across five brands (Hermes, Tesla, IKEA, Patagonia, Erewhon) and cross-model replicated with identical structural findings between Claude and Gemini.
Six Modules
Brand Decomposition
Decompose any brand into a structured signal inventory across all eight dimensions. Each signal is typed by emission source, tagged as designed or ambient, and assessed for strength and reach.
Observer Mapping
Define 3-6 observer cohorts with formal spectral profiles: sensitivity spectrum, dimension weights, tolerances, priors, and identity gate configuration. Perceptual groupings, not demographic segments.
Cloud Prediction
Predict how each observer cohort will cluster the brand's signals into perception clouds. Each cloud has a valence (positive, negative, ambivalent), confidence score, and formation mode (standard, mediated, stalled).
Coherence Audit
Score the brand on seven structural metrics: coverage, gate permeability, cloud coherence, signal strength, re-collapse resistance, emission efficiency, and designed/ambient ratio. Produces a coherence type diagnosis.
Emission Strategy
Design a dimensional emission strategy based on audit findings. Prioritize which dimensions to amplify, reduce, or restructure. Includes action plan with timeline and expected cohort response.
Re-collapse Simulation
Stress-test the brand under disruption scenarios: PR crisis, competitor entry, founder controversy, market shift. Model how each cohort's conviction re-collapses and where the structural vulnerabilities lie.
Supporting Resources
- Atom Taxonomy — machine-readable 8-dimension taxonomy with sub-types, observer template, encounter bundle types
- Framework Specification — complete SBT v2.1 with all mechanisms, discoveries, and metrics
- Glossary — all SBT terms with definitions, cross-references, and traditional-concept mappings
- Framework Reference — established frameworks used per module (ISO 20671, SWOT, RACE, PESTEL)