Psychological Model

PsychAI runs two analysis passes on the same input text:

• Negative markers pass (Master Negative prompt)
• Positive markers pass (Master Positive prompt)

Each pass produces:

• A structured table output (JSON)
• An explanatory narrative output (text)

Negative analysis dimensions

Negative table entries are normalized to:

• Psychological_Marker
• Sub_Type
• Confidence (0-10)
• Valence (-3 to +3)
• Arousal (1-10)
• Prevalence (1-10)

Positive analysis dimensions

Positive table entries are normalized to:

• Skill
• Sub_Dimension
• Competence (0-10)
• Confidence (0-10)
• Skill_Use (Applied, Conceptual, Not Evidenced)
• Reflection_Depth (Low, Moderate, High)

Mode composition

The API exposes nine output modes that return one side (positive/negative) or both sides, as table, explanatory text, or combined payloads.

See Response Schema for exact envelopes by mode.