Reports & AI Narratives

Your Icosa profile is rich with structural data, but numbers alone do not tell a story. Once your profile is computed, the report system turns those 302 metrics into something you can read, share, and use — readable narratives, visual signatures, cross-system translations, and exportable clinical documents.

302 Metrics, One Story

Your Icosa profile contains over 302 individually computed metrics. The computation takes less than a millisecond. Here is what it covers:

• Twenty Harmony scores, four Capacity summaries, five Domain summaries
• A five-layer Coherence score
• Up to nine Gateway evaluations
• Forty-two possible Trap checks and thirty-two Basin scans
• Twenty Fault Line assessments and six Dynamics factors
• Formation classification and band-specific interpretive text for every metric

But no one reads 302 metrics. People read stories.

The Icosa report system translates your structural profile into readable, clinically grounded narrative — then layers on visualizations, cross-system translations, and exportable formats so you can use your results however they serve you best.

AI-Generated Narratives

Every narrative is generated by an AI model (Anthropic Claude) that has been given deep structural knowledge of the Icosa model. The system does not produce generic personality descriptions. It reads your specific profile — your exact grid configuration, your precise Coherence score, which Gateways are open or closed, which Traps are active, what formation you occupy — and writes an interpretation grounded in that unique data.

Three narrative depths are available, matched to assessment tier and user need:

Quick summary (Claude Haiku): A concise overview of your profile shape, your Coherence band, and the most salient features. Best for fast orientation or regular check-ins.

Standard narrative (Claude Sonnet): A thorough section-by-section interpretation covering your capacities, domains, harmonies, coherence, gateways, traps, and formation. This is the default for most users and provides a complete interpretive picture.

Detailed analysis (Claude Opus): A deep, nuanced analysis that explores interactions between grid features, traces causal chains from one part of your profile to another, and surfaces implications that require sustained analytical reasoning to identify. The detailed level is designed for clinicians, researchers, and anyone who wants the most thorough interpretation available.

What the Narrative Covers

Each narrative is structured into sections that build from overview to detail:

• Profile overview: Your formation, Coherence band, and the headline features of your grid
• Capacity analysis: How each of your four Capacities is operating — under, centered, or over — and what that means in practice
• Domain analysis: Where your energy concentrates and where it is thin across the five life Domains
• Harmony deep-dives: The specific intersections where your profile is most notable — your strongest centers, your most strained, and the ones generating the most dynamic tension
• Coherence interpretation: What your five-layer Coherence score means, which layers are contributing and which are suppressing, and what the band placement tells you about your system’s current integration
• Formation description: What your named formation looks like, how it behaves, and what distinguishes it from adjacent formations
• Gateway guidance: Which Gateways are open, which are closed, and why the closed ones matter for your specific configuration
• Trap warnings: Any active Traps with their mechanisms, severity, and escape Gateways
• Centering recommendations: Where movement would have the most impact, based on your unique geometry

If you want to re-read or share, the full narrative is cached and available instantly on subsequent views.

Tone Calibration

The narrative engine adjusts its voice based on who is reading. A self-report for an individual emphasizes personal insight and actionable self-understanding. A clinician report uses clinical vocabulary, includes DSM-adjacent framing, and surfaces risk indicators that would be inappropriate in a consumer-facing narrative. A partner-perspective report acknowledges the observer’s position and the inherent limitations of external assessment.

Demographics also shape tone. The engine adjusts developmental framing, cultural sensitivity, and explanatory depth based on the respondent’s age group, life context, and stated background — not to change the structural findings, but to ensure the interpretation lands accurately for the person reading it.

The Icosaglyph

Your Icosaglyph is the visual signature of your profile: a 4x5 grid that encodes your entire personality architecture in a single image.

Each of the twenty centers is rendered with its current state — centered, under, or over — and colored by Coherence band. Active Traps are highlighted with distinctive markers. Basins are outlined. Gateways are marked. The result is a data-dense visualization that a trained eye can read in seconds, and that anyone can learn to read with a few minutes of orientation.

The Icosaglyph is not decoration. It is a diagnostic tool. Two profiles with the same Coherence score can look dramatically different on the glyph — one showing concentrated distress in the Emotional column, the other showing scattered under-expression across the entire grid. The number is the same; the picture is different; and the clinical implications are different. The glyph captures structural information that a single score cannot.

Icosaglyph variants include grid style (filled centers with color encoding) and line style (outline rendering for Trap and Basin visualization). Both are available as interactive SVG in the browser and as exportable PNG for sharing, printing, or embedding in clinical documentation.

Compare With Seven Systems

If you already think in Big Five, MBTI, Enneagram, or another personality framework, Icosa translates your profile into those familiar terms. Seven systems are supported:

Comparisons are computed structurally, not by asking you a second set of questions. The engine maps your 20-center Icosa profile onto each external system’s theoretical space, producing translations that are grounded in your actual data rather than in self-report to a different instrument.

Three-Layer Hallucination Protection

Any system that translates between personality frameworks risks generating plausible-sounding but structurally unfounded correlations. The Icosa comparison engine addresses this with a three-layer protection system that ensures every comparison claim is grounded in computed data.

Structural gate: Each comparison mapping is defined as a set of explicit mathematical relationships between Icosa grid positions and target-system constructs. The AI narrative generator cannot invent new mappings — it can only describe mappings that the structural engine has computed.

Instruction guard: The narrative prompt explicitly prohibits the AI from generating comparison content that is not supported by the computed structural translation. The model is told what the comparison found; it is not asked to speculate about what it might find.

Postprocess scrubber: After generation, the narrative passes through a scrubbing layer that identifies and removes any comparison claims not supported by the structural computation. If the AI model hallucinated a correlation between an Icosa construct and an external system construct, the scrubber catches it.

This means some comparison sections may be shorter than you expect — that is a feature, not a limitation. Conservative output is trustworthy output.

Structured Reports

Beyond the AI narrative, the report system generates structured data sections: tables, charts, key-value summaries, badge indicators, and progress bars. Twenty-nine section builders produce content that can be rendered in four formats:

• JSON: For programmatic consumption and integration
• HTML: For interactive web display with expandable sections and per-section feedback
• Plain text: For accessibility, copying, and simple sharing
• PDF: For printing, clinical records, and formal documentation

Report presets bundle sections into purposeful packages. A Summary preset gives you the highlights. A Detailed preset includes everything. A Clinical preset adds clinical vocabulary, risk indicators, and DSM-adjacent framing. A SOAP preset formats findings as Subjective-Objective-Assessment-Plan notes for clinical documentation. A Dyadic preset structures the report around relational dynamics. A Timeline preset emphasizes longitudinal change.

Feedback and Accuracy

Every section of the report includes a feedback mechanism. You can rate whether a section accurately describes your experience, and those ratings feed back into the system’s understanding of narrative quality. This is not self-validation theater — it is a genuine accuracy signal. When a section consistently receives low accuracy ratings across many users, that section’s narrative prompt is reviewed and refined.

The system learns which kinds of profiles its narratives handle well and which kinds need improvement. Your feedback contributes directly to that process.

What This Costs

The assessment and basic structural profile are free. AI-generated narratives, framework comparisons, and PDF export are available at the Consumer Premium tier and above. Clinician reports with clinical vocabulary and SOAP formatting are available at the Clinician Starter tier and above.

The structural data — your 302 metrics, your formation, your Coherence score, your Gateways and Traps — is always yours. The narrative layers interpret that data. The comparison layers translate it. The export layers package it. But the foundation is free, and it is complete.